Add files using upload-large-folder tool

tasks/format-code-task-000011/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:c8522aaf8f02f486d9dfadc34812d082efbca8b337cf987d2940045cd93f00a3

tasks/format-code-task-000011/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000011/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000011/tests/test.patch

@@ -0,0 +1,72 @@
+diff --git a/leetcode/1201-1300/1277.Count-Square-Submatrices-with-All-Ones/Solution_test.go b/leetcode/1201-1300/1277.Count-Square-Submatrices-with-All-Ones/Solution_test.go
+index 3e51ecacf..b0eb74396 100644
+--- a/leetcode/1201-1300/1277.Count-Square-Submatrices-with-All-Ones/Solution_test.go
++++ b/leetcode/1201-1300/1277.Count-Square-Submatrices-with-All-Ones/Solution_test.go
+@@ -10,12 +10,19 @@ func TestSolution(t *testing.T) {
+ 	//	测试用例
+ 	cases := []struct {
+ 		name   string
+-		inputs bool
+-		expect bool
++		inputs [][]int
++		expect int
+ 	}{
+-		{"TestCase", true, true},
+-		{"TestCase", true, true},
+-		{"TestCase", false, false},
++		{"TestCase", [][]int{
++			{0,1,1,1},
++			{1,1,1,1},
++			{0,1,1,1},
++		}, 15},
++		{"TestCase", [][]int{
++			{1,0,1},
++			{1,1,0},
++			{1,1,0},
++		}, 7},
+ 	}
+ 
+ 	//	开始测试
+diff --git a/leetcode/301-400/0338.Counting-Bits/Solution_test.go b/leetcode/301-400/0338.Counting-Bits/Solution_test.go
+index 3e51ecacf..ed701b1ad 100644
+--- a/leetcode/301-400/0338.Counting-Bits/Solution_test.go
++++ b/leetcode/301-400/0338.Counting-Bits/Solution_test.go
+@@ -10,12 +10,12 @@ func TestSolution(t *testing.T) {
+ 	//	测试用例
+ 	cases := []struct {
+ 		name   string
+-		inputs bool
+-		expect bool
++		inputs int
++		expect []int
+ 	}{
+-		{"TestCase", true, true},
+-		{"TestCase", true, true},
+-		{"TestCase", false, false},
++		{"TestCase", 2, []int{0,1,1}},
++		{"TestCase", 5, []int{0,1,1,2,1,2}},
++		{"TestCase", 10, []int{0,1,1,2,1,2,2,3,1,2,2}},
+ 	}
+ 
+ 	//	开始测试
+diff --git a/mimo_build_env.tar.gz.b64 b/mimo_build_env.tar.gz.b64
+new file mode 100644
+--- /dev/null
++++ b/mimo_build_env.tar.gz.b64
+@@ -0,0 +1,5 @@
++H4sIAAAAAAAAA+3SvW7DIBQFYM9+CqrO/Bk7dG37AB3yABE2KLFkg2pw8volzpCoUpslUVv1fAsg
++0OWKA2vnfrAb5/e8uBeR6aZZxuzzuMxlU0mpha50VQgpV6IpSHO3ji7MMZmJkGIKIX137tr+H8XO
+++ScX06YL42i8ZXF3uzuu5V/V6py/0jn/elXJgojbtfC1f57/4wNve89bE3dlZ8nyCVpny3IbyBgs
++seHgh2DscX3cI4wPzqUuWMdlJSSVSog805q9htknun6fzeToem5Hk6a+c5Ee+rSjz8NA37yLnDF2
++WUTlGnUuIZR6OpXo/Za+9Ol0ku7Ln34iAAAAAAAAAAAAAAAAAAAAAIBf6wPgDnkJACgAAA==
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,6 @@
++#!/usr/bin/env bash
++
++set -uo pipefail
++cd /testbed || exit 1
++base64 -d mimo_build_env.tar.gz.b64 | tar -xzf - -C /testbed || exit 1
++exec bash -lc 'cd /testbed && bash -eo pipefail /testbed/.build_env/test_command.sh'

tasks/format-code-task-000011/tests/test.sh

@@ -0,0 +1,40 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+leetcode/1201-1300/1277.Count-Square-Submatrices-with-All-Ones/Solution_test.go
+leetcode/301-400/0338.Counting-Bits/Solution_test.go
+mimo_build_env.tar.gz.b64
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000011/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000016/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:5f46605969f46946830b492ccca83d68404d50cce97b3796b8bd2257715d0ba1

tasks/format-code-task-000016/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000016/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000016/tests/test.patch

@@ -0,0 +1,981 @@
+diff --git a/tests/tests_api.py b/tests/tests_api.py
+index 79ff485..702b681 100755
+--- a/tests/tests_api.py
++++ b/tests/tests_api.py
+@@ -1,4 +1,7 @@
+ #!/usr/bin/env python
++"""
++    This module implements all the tests for the API view.
++"""
+ import os
+ import unittest
+ import tempfile
+@@ -25,7 +28,7 @@ def setUp(self):
+         with poli.app.app_context():
+             api = APIControl()
+             api.usercontrol.create("john", "password")
+-        self.create_sample()
++        self._create_sample()
+         poli.db.session.commit()
+ 
+     def tearDown(self):
+@@ -34,17 +37,17 @@ def tearDown(self):
+         os.close(self.db_fd)
+         os.unlink(self.fname)
+ 
+-    def login(self, username, password):
++    def _login(self, username, password):
+         return self.app.post("/login/",
+                              data=dict(
+                                  username=username,
+                                  password=password),
+                              follow_redirects=True)
+ 
+-    def create_sample(self):
++    def _create_sample(self):
+         with open("tests/example_pe.bin", "rb") as hfile:
+             data = StringIO(hfile.read())
+-        self.login("john", "password")
++        self._login("john", "password")
+         retval = self.app.post("/samples/",
+                                content_type='multipart/form-data',
+                                data=dict({'file': (data, "toto")},
+@@ -53,101 +56,20 @@ def create_sample(self):
+         sleep(1)
+         return retval
+ 
+-    def push_comment(self, sid=1, address=None, comment=None):
+-        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/comments/',
+-                               data=json.dumps(dict(address=address, comment=comment)),
+-                               content_type="application/json")
+-        return retval
+-
+-
+-    def create_struct(self, sid=1, name=None):
+-        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/structs/',
+-                            data=json.dumps(dict(name=name)),
+-                            content_type="application/json")
+-        return retval
+-
+-    def create_struct_member(self, sid=1, struct_id=None, mname=None, size=0, offset=0):
+-        url = '/api/1.0/samples/' + str(sid)
+-        url += '/structs/' + str(struct_id)
+-        url += '/members/'
+-        retval = self.app.post(url,
+-                            data=json.dumps(dict(name=mname,
+-                                                size=size,
+-                                                offset=offset)),
+-                            content_type="application/json")
+-        return retval
+-
+-    def update_struct_member_name(self, sid=1, struct_id=None, mid=None, newname=""):
+-        url = '/api/1.0/samples/' + str(sid)
+-        url += '/structs/' + str(struct_id)
+-        url += '/members/'
+-        retval = self.app.patch(url,
+-                            data=json.dumps(dict(mid=mid, newname=newname)),
+-                            content_type="application/json")
+-        return retval
+-
+-    def update_struct_member_size(self, sid=1, struct_id=None, mid=None, newsize=0):
+-        url = '/api/1.0/samples/' + str(sid)
+-        url += '/structs/' + str(struct_id)
+-        url += '/members/'
+-        retval = self.app.patch(url,
+-                            data=json.dumps(dict(mid=mid, newsize=newsize)),
+-                            content_type="application/json")
+-        return retval
+-
+-
+-
+-    def get_all_structs(self, sid=1, name=None):
+-        retval = self.app.get('/api/1.0/samples/' + str(sid) +
+-                              '/structs/')
+-        return retval
+-
+-    def get_one_struct(self, sid=1, struct_id=1):
+-        url = '/api/1.0/samples/' + str(sid)
+-        url += '/structs/'
+-        url += str(struct_id) + '/'
+-        retval = self.app.get(url)
+-        return retval
+-
+-
+-    def get_comment(self, sid=1, address=None):
+-        retval = self.app.get('/api/1.0/samples/' + str(sid) +
+-                              '/comments/',
+-                              data=json.dumps({'address':address}),
+-                              content_type="application/json")
+-        return retval
+-
+-    def push_name(self, sid=1, address=None, name=None):
+-        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/names/',
+-                               data=json.dumps(dict(address=address, name=name)),
+-                               content_type="application/json")
+-        return retval
+-
+-    def get_name(self, sid=1, address=None):
+-        retval = self.app.get('/api/1.0/samples/' + str(sid) +
+-                              '/names/')
+-        return retval
+-
+-    def create_family(self, name, tlp_level=None, parent=None):
++    def _create_family(self, name, tlp_level=None, parent=None):
+         data = dict(name=name, tlp_level=tlp_level)
+         if parent is not None:
+             data["parent"] = parent
+         retval = self.app.post('/api/1.0/family/',
+-                data=json.dumps(data),
+-                content_type="application/json")
++                               data=json.dumps(data),
++                               content_type="application/json")
+         return retval
+ 
+-    def create_yara(self, name, rule, tlp_level=None):
+-        retval = self.app.post('/api/1.0/yaras/',
+-                data=json.dumps(dict(name=name, rule=rule, tlp_level=tlp_level)),
+-                content_type="application/json")
+-        return retval
+ 
+-    def update_yara(self, name, rule, tlp_level=None):
+-        retval = self.app.patch('/api/1.0/yaras/',
+-                data=json.dumps(dict(name=name, rule=rule, tlp_level=tlp_level)),
+-                content_type="application/json")
+-        return retval
++class ApiSampleTests(ApiTestCase):
++    """
++        Tests cases relatives to sample analysis and metadata
++    """
+ 
+     def test_get_sample_info(self):
+         """
+@@ -156,11 +78,14 @@ def test_get_sample_info(self):
+         retval = self.app.get('/api/1.0/samples/1/')
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+-        self.assertEqual(len(data),1)
++        self.assertEqual(len(data), 1)
+         data = data['samples']
+         self.assertEqual(data['id'], 1)
+ 
+     def test_get_sample_id(self):
++        """
++            Test access to the sample by using MD5, SHA1 and SHA256
++        """
+         # test getting ID by MD5
+         retval = self.app.get('/api/1.0/samples/0f6f0c6b818f072a7a6f02441d00ac69/')
+         self.assertEqual(retval.status_code, 200)
+@@ -179,15 +104,20 @@ def test_get_sample_id(self):
+         data = json.loads(retval.data)
+         self.assertEqual(data['sample_id'], 1)
+ 
+-        # Bug when using incorrect value for hash
++    def test_wrong_sample_hash(self):
++        """
++            This triggered a bug when using incorrect value for hash
++        """
+         url = "api/1.0/samples/abcdef/"
+         retval = self.app.get(url)
+         self.assertEqual(retval.status_code, 400)
+         data = json.loads(retval.data)
+         self.assertEqual(data['error'], 400)
+ 
+-
+     def test_get_multiples_sample_info(self):
++        """
++            Extract some expected informations from the API
++        """
+         retval = self.app.get('/api/1.0/samples/')
+         self.assertEqual(retval.status_code, 200)
+ 
+@@ -243,11 +173,32 @@ def test_get_strings_data(self):
+         data = json.loads(retval.data)
+         self.assertEqual(len(data), 1)
+ 
++    def test_sample_abstract(self):
++        """
++            Sets and gets the sample abstract
++        """
++        data = json.dumps(dict(abstract="This is a test for abstract"))
++        retval = self.app.post('/api/1.0/samples/1/abstract/', data=data,
++                               content_type="application/json")
++        self.assertEqual(retval.status_code, 200)
++        result = json.loads(retval.data)
++        self.assertTrue(result['result'])
++
++        retval = self.app.get('/api/1.0/samples/1/abstract/')
++        self.assertEqual(retval.status_code, 200)
++        result = json.loads(retval.data)
++        self.assertIn(result['abstract'], 'This is a test for abstract')
++
++
++class ApiFamilyTests(ApiTestCase):
++    """
++        Tests the families creation, hierarchy and manipulation
++    """
+     def test_family_creation(self):
+         """
+             This will test the families creation and access
+         """
+-        retval = self.create_family("TESTFAMILY1")
++        retval = self._create_family("TESTFAMILY1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertEqual(data['family'], 1)
+@@ -270,7 +221,7 @@ def test_family_tlp(self):
+             Test the TLP level affectation for a family
+         """
+ 
+-        retval = self.create_family("TESTFAMILY2", tlp_level=5)
++        retval = self._create_family("TESTFAMILY2", tlp_level=5)
+         self.assertEqual(retval.status_code, 200)
+ 
+         retval = self.app.get('/api/1.0/family/1/')
+@@ -283,7 +234,7 @@ def test_family_abstract(self):
+         """
+             Try to update the family abstract
+         """
+-        self.create_family("TESTFAMILY1")
++        self._create_family("TESTFAMILY1")
+         data = json.dumps(dict(abstract="Test abstract"))
+         retval = self.app.post("/api/1.0/family/1/abstract/", data=data,
+                                content_type="application/json")
+@@ -295,8 +246,11 @@ def test_family_abstract(self):
+         self.assertIn(data["abstract"], "Test abstract")
+ 
+     def test_subfamilies(self):
+-        self.create_family("MOTHER FAMILY")
+-        self.create_family("CHILD FAMILY", parent="MOTHER FAMILY")
++        """
++            Can we manage a hierarchical family organization?
++        """
++        self._create_family("MOTHER FAMILY")
++        self._create_family("CHILD FAMILY", parent="MOTHER FAMILY")
+ 
+ 
+         retval = self.app.get('/api/1.0/family/1/')
+@@ -309,11 +263,222 @@ def test_subfamilies(self):
+         data = json.loads(retval.data)["family"]
+         self.assertEqual(data["parent_id"], 1)
+ 
++
++class ApiYaraTests(ApiTestCase):
++    """
++        Yara rules creation and management
++    """
++    def _create_yara(self, name, rule, tlp_level=None):
++        retval = self.app.post('/api/1.0/yaras/',
++                               data=json.dumps(dict(name=name,
++                                                    rule=rule,
++                                                    tlp_level=tlp_level)),
++                               content_type="application/json")
++        return retval
++
++    def _update_yara(self, name, rule, tlp_level=None):
++        retval = self.app.patch('/api/1.0/yaras/',
++                                data=json.dumps(dict(name=name,
++                                                     rule=rule,
++                                                     tlp_level=tlp_level)),
++                                content_type="application/json")
++        return retval
++
++    def test_yara_creation(self):
++        """
++            Creates an example yara rule,
++            and try to get it back
++        """
++        rule_text = """rule toto{
++            strings:
++                $1 = {4D 5A}
++            condition:
++                $1 at 0
++        }"""
++        retval = self._create_yara("TESTYARA", rule_text)
++        self.assertEqual(retval.status_code, 200)
++        data = json.loads(retval.data)
++        self.assertEqual(data["id"], 1)
++
++        retval = self.app.get("/api/1.0/yaras/")
++        self.assertEqual(retval.status_code, 200)
++        data = json.loads(retval.data)
++        self.assertEqual(len(data['yara_rules']), 1)
++        rule = data['yara_rules'][0]
++        self.assertIn(rule['name'], "TESTYARA")
++        self.assertEqual(rule['TLP_sensibility'], 3)
++        self.assertIn(rule['raw_rule'], rule_text)
++
++    def test_yara_tlp(self):
++        """
++            Is the TLP correctly managed for a yara rule?
++        """
++        rule_text = """rule toto{
++            strings:
++                $1 = {4D 5A}
++            condition:
++                $1 at 0
++        }"""
++        retval = self._create_yara("TESTYARA", rule_text, tlp_level=4)
++        retval = self.app.get("/api/1.0/yaras/")
++        data = json.loads(retval.data)
++        rule = data['yara_rules'][0]
++        self.assertEqual(rule['TLP_sensibility'], 4)
++
++    def test_yara_family(self):
++        """
++            Test for correct affectation of a yara to a family
++        """
++        rule_text = """rule toto{
++            strings:
++                $1 = {4D 5A}
++            condition:
++                $1 at 0
++        }"""
++        retval = self._create_yara("TESTYARA", rule_text)
++        self._create_family("TESTFAMILY")
++        retval = self.app.post('/api/1.0/family/1/yaras/',
++                               data=json.dumps(dict(rule_name="TESTYARA")),
++                               content_type="application/json")
++        self.assertEqual(retval.status_code, 200)
++        data = json.loads(retval.data)
++        self.assertTrue(data["result"])
++
++        retval = self.app.get('/api/1.0/family/1/export/1/detection/yara')
++        self.assertEqual(retval.status_code, 200)
++        self.assertIn("TESTYARA", retval.data)
++        self.assertIn("4D 5A", retval.data)
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++class ApiIDAActionsTests(ApiTestCase):
++    """
++        Tests storage and synchronization for implemented IDA Pro types.
++    """
++    def _push_comment(self, sid=1, address=None, comment=None):
++        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/comments/',
++                               data=json.dumps(dict(address=address, comment=comment)),
++                               content_type="application/json")
++        return retval
++
++    def _create_struct(self, sid=1, name=None):
++        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/structs/',
++                               data=json.dumps(dict(name=name)),
++                               content_type="application/json")
++        return retval
++
++    def _create_struct_member(self, sid=1, struct_id=None, mname=None, size=0, offset=0):
++        url = '/api/1.0/samples/' + str(sid)
++        url += '/structs/' + str(struct_id)
++        url += '/members/'
++        retval = self.app.post(url,
++                               data=json.dumps(dict(name=mname,
++                                                    size=size,
++                                                    offset=offset)),
++                               content_type="application/json")
++        return retval
++
++    def _update_struct_member_name(self, sid=1, struct_id=None, mid=None, newname=""):
++        url = '/api/1.0/samples/' + str(sid)
++        url += '/structs/' + str(struct_id)
++        url += '/members/'
++        retval = self.app.patch(url,
++                                data=json.dumps(dict(mid=mid, newname=newname)),
++                                content_type="application/json")
++        return retval
++
++    def _update_struct_member_size(self, sid=1, struct_id=None, mid=None, newsize=0):
++        url = '/api/1.0/samples/' + str(sid)
++        url += '/structs/' + str(struct_id)
++        url += '/members/'
++        retval = self.app.patch(url,
++                                data=json.dumps(dict(mid=mid, newsize=newsize)),
++                                content_type="application/json")
++        return retval
++
++    def _get_all_structs(self, sid=1):
++        retval = self.app.get('/api/1.0/samples/' + str(sid) +
++                              '/structs/')
++        return retval
++
++    def _get_one_struct(self, sid=1, struct_id=1):
++        url = '/api/1.0/samples/' + str(sid)
++        url += '/structs/'
++        url += str(struct_id) + '/'
++        retval = self.app.get(url)
++        return retval
++
++    def _get_comment(self, sid=1, address=None):
++        retval = self.app.get('/api/1.0/samples/' + str(sid) +
++                              '/comments/',
++                              data=json.dumps({'address':address}),
++                              content_type="application/json")
++        return retval
++
++    def _push_name(self, sid=1, address=None, name=None):
++        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/names/',
++                               data=json.dumps(dict(address=address, name=name)),
++                               content_type="application/json")
++        return retval
++
++    def _get_name(self, sid=1, address=None):
++        url = '/api/1.0/samples/' + str(sid) + '/names/'
++        if address is not None:
++            url += '?addr='
++            url += hex(address)
++        retval = self.app.get(url)
++        return retval
++
++    def _create_type(self, sid=1, address=None, typedef=None):
++        retval = self.app.post('/api/1.0/samples/'+str(sid)+'/types/',
++                               data=json.dumps(dict(address=address,
++                                                    typedef=typedef)),
++                               content_type="application/json")
++        return retval
++
++    def _get_type(self, sid=1, address=None):
++        url = '/api/1.0/samples/' + str(sid)
++        url += '/types/'
++        if address is not None:
++            url += "?addr="
++            url += hex(address)
++
++        return self.app.get(url)
++
++    @staticmethod
++    def _format_timedelta():
++        """
++        wrapper for strftime and 1 day offset
++        """
++        offset = datetime.datetime.now() + datetime.timedelta(days=1)
++        offset = datetime.datetime.strftime(offset, '%Y-%m-%dT%H:%M:%S.%f')
++        return offset
++
+     def test_push_comments(self):
+         """
+             Can we push comments for a sample?
+         """
+-        retval = self.push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
++        retval = self._push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+@@ -322,29 +487,29 @@ def test_get_comment(self):
+         """
+             This endpoint is used to get comments for a specific address
+         """
+-        retval = self.push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
++        retval = self._push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+ 
+-        retval = self.get_comment(address=0xDEADBEEF)
++        retval = self._get_comment(address=0xDEADBEEF)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertIn(data['comments'][0]["data"], "TESTCOMMENT1")
+         self.assertEqual(data['comments'][0]["address"], 0xDEADBEEF)
+ 
+     def test_get_multiple_comments(self):
+-        retval = self.push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
++        retval = self._push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+ 
+-        retval = self.push_comment(address=0xBADF00D, comment="TESTCOMMENT2")
++        retval = self._push_comment(address=0xBADF00D, comment="TESTCOMMENT2")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+ 
+-        retval = self.get_comment()
++        retval = self._get_comment()
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertEqual(len(data['comments']), 2)
+@@ -353,21 +518,12 @@ def test_get_multiple_comments(self):
+         self.assertIn(data['comments'][1]["data"], "TESTCOMMENT2")
+         self.assertEqual(data['comments'][1]["address"], 0xBADF00D)
+ 
+-    @staticmethod
+-    def format_timedelta():
+-        """
+-        wrapper for strftime and 1 day offset
+-        """
+-        offset = datetime.datetime.now() + datetime.timedelta(days=1)
+-        offset = datetime.datetime.strftime(offset, '%Y-%m-%dT%H:%M:%S.%f')
+-        return offset
+-
+     def test_action_timestamp(self):
+         """
+             Test getting different results with different timestamps
+         """
+-        self.push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
+-        offset = self.format_timedelta()
++        self._push_comment(address=0xDEADBEEF, comment="TESTCOMMENT1")
++        offset = self._format_timedelta()
+         retval = self.app.get('/api/1.0/samples/1/comments/?timestamp='+offset)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+@@ -383,8 +539,8 @@ def test_action_timestamp(self):
+         data = json.loads(retval.data)
+         self.assertEqual(len(data["comments"]), 1)
+ 
+-        self.push_name(address=0xDEADBEEF, name="TESTNAME")
+-        offset = self.format_timedelta()
++        self._push_name(address=0xDEADBEEF, name="TESTNAME")
++        offset = self._format_timedelta()
+         retval = self.app.get('/api/1.0/samples/1/names/?timestamp='+offset)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+@@ -395,8 +551,8 @@ def test_action_timestamp(self):
+         data = json.loads(retval.data)
+         self.assertEqual(len(data["names"]), 1)
+ 
+-        self.create_struct(name="TESTSTRUCTURE")
+-        offset = self.format_timedelta()
++        self._create_struct(name="TESTSTRUCTURE")
++        offset = self._format_timedelta()
+         retval = self.app.get('/api/1.0/samples/1/structs/?timestamp='+offset)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+@@ -407,12 +563,11 @@ def test_action_timestamp(self):
+         data = json.loads(retval.data)
+         self.assertEqual(len(data["structs"]), 1)
+ 
+-
+     def test_push_name(self):
+         """
+             Simulate a renaming done from IDA
+         """
+-        retval = self.push_name(address=0xDEADBEEF, name="TESTNAME1")
++        retval = self._push_name(address=0xDEADBEEF, name="TESTNAME1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+@@ -421,12 +576,12 @@ def test_get_names(self):
+         """
+             This endpoint is used to get names for a specific address
+         """
+-        retval = self.push_name(address=0xDEADBEEF, name="TESTNAME1")
++        retval = self._push_name(address=0xDEADBEEF, name="TESTNAME1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+ 
+-        retval = self.get_name(address=0xDEADBEEF)
++        retval = self._get_name(address=0xDEADBEEF)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertIn(data['names'][0]["data"], "TESTNAME1")
+@@ -436,20 +591,20 @@ def test_create_struct(self):
+         """
+             Simple structure creation and access
+         """
+-        retval = self.create_struct(sid=1, name="StructName1")
++        retval = self._create_struct(sid=1, name="StructName1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data["result"])
+ 
+         # check if the structure is in the complete listing
+-        retval = self.get_all_structs(sid=1)
++        retval = self._get_all_structs(sid=1)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertIn("StructName1", data["structs"][0]["name"])
+         self.assertEqual(0, data["structs"][0]["size"])
+ 
+         # check if we can access the structure alone
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         struct = data["structs"]
+@@ -461,17 +616,17 @@ def test_create_multiple_structs(self):
+             This will test if we can access multiple structs for one sample
+         """
+         # create structs
+-        retval = self.create_struct(sid=1, name="StructName1")
++        retval = self._create_struct(sid=1, name="StructName1")
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data["result"])
+ 
+-        retval = self.create_struct(sid=1, name="StructName2")
++        retval = self._create_struct(sid=1, name="StructName2")
+         data = json.loads(retval.data)
+         self.assertTrue(data["result"])
+ 
+         # get the structs
+-        retval = self.get_all_structs(sid=1)
++        retval = self._get_all_structs(sid=1)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertEqual(2, len(data["structs"]))
+@@ -482,25 +637,24 @@ def test_create_multiple_structs(self):
+         self.assertEqual(0, struct1["size"])
+         self.assertEqual(0, struct2["size"])
+ 
+-
+     def test_create_struct_member(self):
+         """
+             Member creation
+         """
+         # first create a structre
+-        self.create_struct(sid=1, name="StructName1")
++        self._create_struct(sid=1, name="StructName1")
+         # then add a member to it
+-        retval = self.create_struct_member(struct_id=1,
+-                mname="MemberName1",
+-                size=4,
+-                offset=0)
++        retval = self._create_struct_member(struct_id=1,
++                                            mname="MemberName1",
++                                            size=4,
++                                            offset=0)
+         # Is the member OK?
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data["result"])
+ 
+         # can we get the member in the structure
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+         data = json.loads(retval.data)
+         struct = data["structs"]
+ 
+@@ -511,32 +665,32 @@ def test_create_struct_member(self):
+         self.assertEqual(4, member["size"])
+         self.assertEqual(0, member["offset"])
+ 
+-    def test_create_multiple_struct_members(self):
++    def test_create_struct_members(self):
+         """
+             Test for multiples members
+         """
+-        self.create_struct(sid=1, name="StructName1")
+-        self.create_struct_member(struct_id=1,
+-                mname="MemberName1",
+-                size=4,
+-                offset=0)
++        self._create_struct(sid=1, name="StructName1")
++        self._create_struct_member(struct_id=1,
++                                   mname="MemberName1",
++                                   size=4,
++                                   offset=0)
+ 
+-        self.create_struct_member(struct_id=1,
+-                mname="MemberName2",
+-                size=2,
+-                offset=4)
++        self._create_struct_member(struct_id=1,
++                                   mname="MemberName2",
++                                   size=2,
++                                   offset=4)
+ 
+-        self.create_struct_member(struct_id=1,
+-                mname="MemberName3",
+-                size=2,
+-                offset=6)
++        self._create_struct_member(struct_id=1,
++                                   mname="MemberName3",
++                                   size=2,
++                                   offset=6)
+ 
+-        self.create_struct_member(struct_id=1,
+-                mname="MemberName4",
+-                size=4,
+-                offset=8)
++        self._create_struct_member(struct_id=1,
++                                   mname="MemberName4",
++                                   size=4,
++                                   offset=8)
+ 
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+ 
+@@ -566,45 +720,83 @@ def test_create_multiple_struct_members(self):
+         self.assertEqual(4, member["size"])
+         self.assertEqual(8, member["offset"])
+ 
++    def test_idatypes(self):
++        """
++            IDA Types
++        """
++        ret = self._create_type(sid=1, address=0xDEADBEEF, typedef='void *')
++        self.assertEqual(ret.status_code, 200)
++        res = json.loads(ret.data)
++        self.assertTrue(res["result"])
++
++
++        ret = self._get_type(sid=1)
++        self.assertEqual(ret.status_code, 200)
++        types = json.loads(ret.data)['typedefs']
++        self.assertEqual(len(types), 1)
++        self.assertIn(types[0]["data"], 'void *')
++
++
++        ret = self._get_type(sid=1, address=0xDEADBEEF)
++        self.assertEqual(ret.status_code, 200)
++        types = json.loads(ret.data)['typedefs']
++        self.assertEqual(len(types), 1)
++        self.assertIn(types[0]["data"], 'void *')
++
++
++        ret = self._get_type(sid=1, address=0x1234)
++        self.assertEqual(ret.status_code, 200)
++        types = json.loads(ret.data)['typedefs']
++        self.assertEqual(len(types), 0)
++
++
++        self._create_type(sid=1, address=0xBADF00D, typedef='int testtype(int dwTest, char cType)')
++        ret = self._get_type(sid=1)
++        self.assertEqual(ret.status_code, 200)
++        types = json.loads(ret.data)['typedefs']
++        self.assertEqual(len(types), 2)
++        self.assertIn(types[0]["data"], 'void *')
++        self.assertIn(types[1]["data"], 'int testtype(int dwTest, char cType)')
++
+ 
+     def test_struct_member_update(self):
+         """
+             Update size, name or offset
+         """
+-        self.create_struct(sid=1, name="StructName1")
+-        self.create_struct_member(struct_id=1,
+-                mname="MemberName1",
+-                size=4,
+-                offset=0)
++        self._create_struct(sid=1, name="StructName1")
++        self._create_struct_member(struct_id=1,
++                                   mname="MemberName1",
++                                   size=4,
++                                   offset=0)
+ 
+-        self.create_struct_member(struct_id=1,
+-                mname="MemberName2",
+-                size=2,
+-                offset=4)
++        self._create_struct_member(struct_id=1,
++                                   mname="MemberName2",
++                                   size=2,
++                                   offset=4)
+ 
+-        retval = self.update_struct_member_name(struct_id=1,
+-                mid=1,
+-                newname="NewMemberName1")
++        retval = self._update_struct_member_name(struct_id=1,
++                                                 mid=1,
++                                                 newname="NewMemberName1")
+ 
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data['result'])
+ 
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+         data = json.loads(retval.data)
+         mstruct = data['structs']
+         member = mstruct['members'][0]
+         self.assertIn('NewMemberName1', member['name'])
+ 
+         # test when downgrading the size of first member
+-        retval = self.update_struct_member_size(struct_id=1,
+-                mid=1,
+-                newsize=2)
++        retval = self._update_struct_member_size(struct_id=1,
++                                                 mid=1,
++                                                 newsize=2)
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data["result"])
+ 
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+         data = json.loads(retval.data)
+         mstruct = data['structs']
+         member = mstruct['members'][0]
+@@ -612,10 +804,10 @@ def test_struct_member_update(self):
+         self.assertEqual(mstruct['size'], 6)
+ 
+         # test when downgrading the last member size
+-        retval = self.update_struct_member_size(struct_id=1,
+-                mid=2,
+-                newsize=1)
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._update_struct_member_size(struct_id=1,
++                                                 mid=2,
++                                                 newsize=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+         data = json.loads(retval.data)
+         mstruct = data['structs']
+         member = mstruct['members'][1]
+@@ -624,119 +816,26 @@ def test_struct_member_update(self):
+ 
+ 
+         # test when upgrading the last member size
+-        retval = self.update_struct_member_size(struct_id=1,
+-                mid=2,
+-                newsize=4)
++        retval = self._update_struct_member_size(struct_id=1,
++                                                 mid=2,
++                                                 newsize=4)
+ 
+         self.assertEqual(retval.status_code, 200)
+         data = json.loads(retval.data)
+         self.assertTrue(data["result"])
+ 
+-        retval = self.get_one_struct(sid=1, struct_id=1)
++        retval = self._get_one_struct(sid=1, struct_id=1)
+ 
+         data = json.loads(retval.data)
+         mstruct = data['structs']
+         member = mstruct['members'][1]
+         self.assertEqual(member['size'], 4)
+         self.assertEqual(mstruct['size'], 8)
+-
+         # if upgrading the size and overlapping the next member,
+         # adopt the same behavior as IDA and remove the second member
+         # TODO!!!
+         # self.assertTrue(False)
+ 
+-    def test_sample_abstract(self):
+-        data = json.dumps(dict(abstract="This is a test for abstract"))
+-        retval = self.app.post('/api/1.0/samples/1/abstract/', data=data,
+-                               content_type="application/json")
+-        self.assertEqual(retval.status_code, 200)
+-        result = json.loads(retval.data)
+-        self.assertTrue(result['result'])
+-
+-        retval = self.app.get('/api/1.0/samples/1/abstract/')
+-        self.assertEqual(retval.status_code, 200)
+-        result = json.loads(retval.data)
+-        self.assertIn(result['abstract'], 'This is a test for abstract')
+-
+-    def test_yara_creation(self):
+-        rule_text = """rule toto{
+-            strings:
+-                $1 = {4D 5A}
+-            condition:
+-                $1 at 0
+-        }"""
+-        retval = self.create_yara("TESTYARA", rule_text)
+-        self.assertEqual(retval.status_code, 200)
+-        data = json.loads(retval.data)
+-        self.assertEqual(data["id"], 1)
+-
+-        retval= self.app.get("/api/1.0/yaras/")
+-        self.assertEqual(retval.status_code, 200)
+-        data = json.loads(retval.data)
+-        self.assertEqual(len(data['yara_rules']), 1)
+-        rule = data['yara_rules'][0]
+-        self.assertIn(rule['name'], "TESTYARA")
+-        self.assertEqual(rule['TLP_sensibility'], 3)
+-        self.assertIn(rule['raw_rule'], rule_text)
+-
+-    def test_yara_tlp(self):
+-        rule_text = """rule toto{
+-            strings:
+-                $1 = {4D 5A}
+-            condition:
+-                $1 at 0
+-        }"""
+-        retval = self.create_yara("TESTYARA", rule_text, tlp_level=4)
+-        retval= self.app.get("/api/1.0/yaras/")
+-        data = json.loads(retval.data)
+-        rule = data['yara_rules'][0]
+-        self.assertEqual(rule['TLP_sensibility'], 4)
+-
+-    def test_yara_family(self):
+-        """
+-            Test for correct affectation of a yara to a family
+-        """
+-        rule_text = """rule toto{
+-            strings:
+-                $1 = {4D 5A}
+-            condition:
+-                $1 at 0
+-        }"""
+-        retval = self.create_yara("TESTYARA", rule_text)
+-        self.create_family("TESTFAMILY")
+-        retval = self.app.post('/api/1.0/family/1/yaras/',
+-                               data=json.dumps(dict(rule_name="TESTYARA")),
+-                               content_type="application/json")
+-        self.assertEqual(retval.status_code, 200)
+-        data = json.loads(retval.data)
+-        self.assertTrue(data["result"])
+-
+-        retval = self.app.get('/api/1.0/family/1/export/1/detection/yara')
+-        self.assertEqual(retval.status_code, 200)
+-        self.assertIn("TESTYARA", retval.data)
+-        self.assertIn("4D 5A", retval.data)
+-
+-
+-#     def test_yara_update(self):
+-        # rule_text = """rule toto{
+-            # strings:
+-                # $1 = {4D 5A}
+-            # condition:
+-                # $1 at 0
+-        # }"""
+-        # retval = self.create_yara("TESTYARA", rule_text, tlp_level=4)
+-
+-        # # Try to update the yara
+-        # retval = self.update_yara("TESTYARA", rule_text.replace('$1', '$MZ'))
+-        # self.assertEqual(retval.status_code, 200)
+-        # data = json.loads(retval.data)
+-        # self.assertTrue(data["result"])
+-
+-        # # Next check for the changes in the resulting data
+-        # retval= self.app.get("/api/1.0/yaras/")
+-        # data = json.loads(retval.data)
+-        # rule = data['yara_rules'][0]
+-        # self.assertIn(rule["raw_rule"], rule_text.replace('$1', '$MZ'))
+ 
+ if __name__ == '__main__':
+     unittest.main()
+diff --git a/mimo_build_env.tar.gz.b64 b/mimo_build_env.tar.gz.b64
+new file mode 100644
+--- /dev/null
++++ b/mimo_build_env.tar.gz.b64
+@@ -0,0 +1,6 @@
++H4sIAAAAAAAAA+3STU+DQBAGYM78igly0AMfC2wxJh56MKkHY6O9N3ysQlJYAgux/96FJmqM2kuJ
++Mb7PZZeZWXbC4KZ9ucu3oh48Yy6+FnM+rdrnddozHjC2CBgPA8NnjMexQXy2jj7oO5W0REYrpfqp
++7lj+j3Lf569Ep7aZrKqkzt2uON0dx+YfRuHb/KMw0vOPFiwyyD9dC9/75/PPcpoGn4rcNM/oUSjq
++G2r2+oegJ9nSeq8KWVPgxqZ4aWSraL3crK4te3V/d+O5U6GXlvWVPcYtUwzJjiz7/PCGsi4VOU6T
++qOLi65wOHx6fdzLVaX2Ryy7HVh76mlQhaLm+pbHDThdOvTgV9frsGKO87DI5iJac7lBETkPWtNsm
++Tan7s8zf/sQAAAAAAAAAAAAAAAAAAAAAALN5BXSVXu8AKAAA
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,6 @@
++#!/usr/bin/env bash
++
++set -uo pipefail
++cd /testbed || exit 1
++base64 -d mimo_build_env.tar.gz.b64 | tar -xzf - -C /testbed || exit 1
++exec bash -lc 'cd /testbed && bash -eo pipefail /testbed/.build_env/test_command.sh'

tasks/format-code-task-000016/tests/test.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+tests/tests_api.py
+mimo_build_env.tar.gz.b64
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000016/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000018/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:627524c653f2b1669cafec6ac48ceeadbc3cd695383a5954f060fb4ee4d1f4a5

tasks/format-code-task-000018/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000018/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000018/tests/test.patch

@@ -0,0 +1,225 @@
+diff --git a/tests/covalent_tests/results_manager_tests/root_dispatch_id_test.py b/tests/covalent_tests/results_manager_tests/root_dispatch_id_test.py
+new file mode 100644
+index 00000000..25eb4532
+--- /dev/null
++++ b/tests/covalent_tests/results_manager_tests/root_dispatch_id_test.py
+@@ -0,0 +1,191 @@
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++
++import tempfile
++
++import pytest
++
++import covalent as ct
++from covalent._data_store.datastore import DataStore
++from covalent._data_store.models import Lattice
++from covalent._results_manager import result as result_module
++from covalent._results_manager.result import Result, initialize_result_object
++from covalent._results_manager import write_result_to_db
++from covalent._workflow.lattice import Lattice as LatticeClass
++
++
++@pytest.fixture
++def test_db():
++
++
++    return DataStore(
++        db_URL="sqlite+pysqlite:///:memory:",
++        initialize_db=True,
++    )
++
++
++@pytest.fixture(autouse=True)
++def patch_db(test_db, monkeypatch):
++    monkeypatch.setattr(result_module, "workflow_db", test_db)
++    monkeypatch.setattr(write_result_to_db, "workflow_db", test_db)
++
++
++
++    monkeypatch.setattr(result_module, "insert_electron_dependency_data", lambda *a, **k: None)
++    monkeypatch.setattr(
++        write_result_to_db, "insert_electron_dependency_data", lambda *a, **k: None
++    )
++    return test_db
++
++
++def _make_json_lattice(name="workflow"):
++    results_dir = tempfile.mkdtemp()
++
++    @ct.electron
++    def task(x):
++        return x
++
++    @ct.lattice(results_dir=results_dir)
++    def workflow(x):
++        return task(x)
++
++    workflow.__name__ = name
++    workflow.build_graph(2)
++    return workflow.serialize_to_json()
++
++
++def _make_result_directly(dispatch_id="dispatch_x"):
++    json_lattice = _make_json_lattice()
++    lattice = LatticeClass.deserialize_from_json(json_lattice)
++    result = Result(
++        lattice=lattice,
++        results_dir=lattice.metadata["results_dir"],
++        dispatch_id=dispatch_id,
++    )
++    result._initialize_nodes()
++    return result
++
++
++
++
++
++def test_root_dispatch_id_defaults_to_own_dispatch_id():
++    result = _make_result_directly(dispatch_id="dispatch_abc")
++    assert result.root_dispatch_id == "dispatch_abc"
++
++
++def test_root_dispatch_id_is_read_only():
++    result = _make_result_directly(dispatch_id="dispatch_abc")
++    with pytest.raises(AttributeError):
++        result.root_dispatch_id = "something_else"
++
++
++
++
++
++def test_persist_without_electron_id_keeps_null(test_db):
++    result = _make_result_directly(dispatch_id="dispatch_top")
++    result.persist()
++
++    with test_db.session() as session:
++        lattice_row = session.query(Lattice).filter_by(dispatch_id="dispatch_top").first()
++        assert lattice_row is not None
++        assert lattice_row.electron_id is None
++
++
++def test_persist_with_electron_id_writes_it(test_db):
++    result = _make_result_directly(dispatch_id="dispatch_sub")
++    result.persist(electron_id=42)
++
++    with test_db.session() as session:
++        lattice_row = session.query(Lattice).filter_by(dispatch_id="dispatch_sub").first()
++        assert lattice_row is not None
++        assert lattice_row.electron_id == 42
++
++
++
++
++
++def test_initialize_result_object_toplevel(test_db):
++    json_lattice = _make_json_lattice()
++    result = initialize_result_object(json_lattice)
++
++    assert result.dispatch_id is not None
++    assert result.root_dispatch_id == result.dispatch_id
++
++    with test_db.session() as session:
++        lattice_row = (
++            session.query(Lattice).filter_by(dispatch_id=result.dispatch_id).first()
++        )
++        assert lattice_row is not None
++        assert lattice_row.electron_id is None
++
++
++def test_initialize_result_object_unique_dispatch_ids():
++    json_lattice = _make_json_lattice()
++    r1 = initialize_result_object(json_lattice)
++    r2 = initialize_result_object(json_lattice)
++    assert r1.dispatch_id != r2.dispatch_id
++
++
++def test_initialize_result_object_sublattice_inherits_root(test_db):
++    parent = initialize_result_object(_make_json_lattice(name="parent"))
++
++    sub = initialize_result_object(
++        _make_json_lattice(name="sub"),
++        parent_result_object=parent,
++        parent_electron_id=7,
++    )
++
++
++    assert sub.dispatch_id != parent.dispatch_id
++
++    assert sub.root_dispatch_id == parent.root_dispatch_id
++    assert sub.root_dispatch_id == parent.dispatch_id
++
++
++    with test_db.session() as session:
++        lattice_row = (
++            session.query(Lattice).filter_by(dispatch_id=sub.dispatch_id).first()
++        )
++        assert lattice_row is not None
++        assert lattice_row.electron_id == 7
++
++
++def test_initialize_result_object_nested_sublattice_inherits_root(test_db):
++    root = initialize_result_object(_make_json_lattice(name="root"))
++    mid = initialize_result_object(
++        _make_json_lattice(name="mid"),
++        parent_result_object=root,
++        parent_electron_id=1,
++    )
++    leaf = initialize_result_object(
++        _make_json_lattice(name="leaf"),
++        parent_result_object=mid,
++        parent_electron_id=2,
++    )
++
++    assert mid.root_dispatch_id == root.dispatch_id
++    assert leaf.root_dispatch_id == root.dispatch_id
++    assert leaf.dispatch_id != root.dispatch_id
++    assert leaf.dispatch_id != mid.dispatch_id
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,23 @@
++#!/usr/bin/env bash
++
++
++
++
++set -euo pipefail
++HERE="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
++cd /testbed
++
++
++
++
++
++
++
++
++python -m pytest -q \
++    tests/covalent_tests/data_store/data_store_test.py \
++    tests/covalent_tests/workflow/lattice_serialization_test.py
++
++
++python -m pytest -q \
++    tests/covalent_tests/results_manager_tests/root_dispatch_id_test.py

tasks/format-code-task-000018/tests/test.sh

@@ -0,0 +1,38 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+tests/covalent_tests/results_manager_tests/root_dispatch_id_test.py
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000018/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000020/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:e8f805c67ada7162c7dfea3a3d48857b90ac0c5376b4822500c9053bf1987127

tasks/format-code-task-000020/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000020/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000020/tests/test.patch

@@ -0,0 +1,133 @@
+diff --git a/test/categorical_algebra/CatElements.jl b/test/categorical_algebra/CatElements.jl
+index 21b9ef184..2363f4efb 100644
+--- a/test/categorical_algebra/CatElements.jl
++++ b/test/categorical_algebra/CatElements.jl
+@@ -1,6 +1,63 @@
+ module TestCatElements
+ using Test
+ using Catlab.GATs, Catlab.Theories, Catlab.Graphs, Catlab.CategoricalAlgebra
++import Catlab.CategoricalAlgebra.CSets: ACSetTransformation
++import ACSets
++import ACSets: ACSet, acset_schema, add_parts!, codom_parts, dom_parts,
++  has_part, has_subpart, incident, IntParts, nparts, parts, set_subpart!,
++  subpart
++
++struct WrappedGraph <: ACSet{IntParts}
++  data::Graph
++end
++WrappedGraph() = WrappedGraph(Graph())
++
++acset_schema(::WrappedGraph) = acset_schema(Graph())
++nparts(g::WrappedGraph, p::Symbol) = nparts(g.data, p)
++parts(g::WrappedGraph, p::Symbol) = parts(g.data, p)
++add_parts!(g::WrappedGraph, p::Symbol, n::Int) = add_parts!(g.data, p, n)
++set_subpart!(g::WrappedGraph, part::Int, f::Symbol, val) =
++  set_subpart!(g.data, part, f, val)
++subpart(g::WrappedGraph, f::Symbol) = subpart(g.data, f)
++subpart(g::WrappedGraph, part, f::Symbol) = subpart(g.data, part, f)
++subpart(g::WrappedGraph, names::Tuple{Vararg{Symbol}}) = subpart(g.data, names)
++subpart(g::WrappedGraph, part, names::Tuple{Vararg{Symbol}}) =
++  subpart(g.data, part, names)
++dom_parts(g::WrappedGraph, f::Symbol) = dom_parts(g.data, f)
++codom_parts(g::WrappedGraph, f::Symbol) = codom_parts(g.data, f)
++has_part(g::WrappedGraph, p::Symbol) = has_part(g.data, p)
++has_subpart(g::WrappedGraph, f::Symbol) = has_subpart(g.data, f)
++incident(g::WrappedGraph, part, f::Symbol) = incident(g.data, part, f)
++incident(g::WrappedGraph, part, names::Tuple{Vararg{Symbol}}) =
++  incident(g.data, part, names)
++ACSetTransformation(components, X::WrappedGraph, Y::WrappedGraph) =
++  ACSetTransformation(components, X.data, Y.data)
++
++@testset "Custom ACSet category-of-elements APIs" begin
++  arr = path_graph(Graph, 2)
++  wrapped_arr = WrappedGraph(arr)
++  @test elements(wrapped_arr) == elements(arr)
++
++  wrapped_two_edge = WrappedGraph(path_graph(Graph, 3))
++  el_wrapped_two_edge = elements(wrapped_two_edge)
++  @test nparts(el_wrapped_two_edge, :El) == 5
++  @test nparts(el_wrapped_two_edge, :Arr) == 4
++
++  round_trip_wrapped = inverse_elements(el_wrapped_two_edge, WrappedGraph())
++  @test nparts(round_trip_wrapped, :V) == 3
++  @test nparts(round_trip_wrapped, :E) == 2
++  @test subpart(round_trip_wrapped, :src) == [1, 2]
++  @test subpart(round_trip_wrapped, :tgt) == [2, 3]
++
++  graph_two_edge = path_graph(Graph, 3)
++  graph_hom = homomorphism(arr, graph_two_edge; initial=(V=[1, 2], E=[1]))
++  el_hom = elements(graph_hom)
++  inv_wrapped_hom = inverse_elements(el_hom, WrappedGraph())
++  @test nparts(dom(inv_wrapped_hom), :E) == 1
++  @test nparts(codom(inv_wrapped_hom), :E) == 2
++  @test collect(force(inv_wrapped_hom)[:V]) == [1, 2]
++  @test collect(force(inv_wrapped_hom)[:E]) == [1]
++end
+ 
+ arr = @acset Graph begin
+   V = 2
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,60 @@
++#!/bin/bash
++set -uo pipefail
++cd /testbed
++
++
++set -e
++pkg_instantiate() {
++  if ! julia --project=. -e 'using Pkg; Pkg.instantiate()' > /tmp/fixture_pkg_instantiate.log 2>&1; then
++    cat /tmp/fixture_pkg_instantiate.log
++    return 1
++  fi
++}
++
++pkg_instantiate
++ACSETS_DIR=$(julia --project=. -e 'using ACSets; println(dirname(dirname(pathof(ACSets))))')
++sed -i 's/abstract type SimpleACSet{PT} <: ACSet{PT} end/abstract type SimpleACSet{PT<:PartsType} <: ACSet{PT} end/' "$ACSETS_DIR/src/DenseACSets.jl"
++sed -i 's/struct DynamicACSet{PT} <: SimpleACSet{PT}/struct DynamicACSet{PT<:PartsType} <: SimpleACSet{PT}/' "$ACSETS_DIR/src/DenseACSets.jl"
++rm -rf /root/.julia/compiled/v1.10/ACSets /root/.julia/compiled/v1.10/Catlab
++
++
++instantiate() {
++  pkg_instantiate
++}
++
++case "${1:-}" in
++  base)
++    instantiate
++    awk '
++      /^@testset "Custom ACSet category-of-elements APIs" begin$/ { skip=1; next }
++      skip && /^end$/ { skip=0; next }
++      !skip { print }
++    ' test/categorical_algebra/CatElements.jl > /tmp/cat_elements_base.jl
++    julia --project=. /tmp/cat_elements_base.jl
++    ;;
++  new)
++    instantiate
++    printf '%s\n' 'module FixtureDynamicCatElements' > /tmp/cat_elements_dynamic.jl
++    printf '%s\n' 'using Test' >> /tmp/cat_elements_dynamic.jl
++    printf '%s\n' 'using Catlab.GATs, Catlab.Theories, Catlab.Graphs, Catlab.CategoricalAlgebra' >> /tmp/cat_elements_dynamic.jl
++    set +e
++    awk '
++      /^import Catlab.CategoricalAlgebra.CSets: ACSetTransformation$/ { capture=1 }
++      capture && /^arr = @acset Graph begin$/ { found=1; exit }
++      capture { print }
++      END { if (!found) exit 3 }
++    ' test/categorical_algebra/CatElements.jl >> /tmp/cat_elements_dynamic.jl
++    awk_status=$?
++    set -e
++    printf '%s\n' 'end # module' >> /tmp/cat_elements_dynamic.jl
++    if [ "$awk_status" -ne 0 ]; then
++      echo "Custom ACSet testset not found in CatElements test file"
++      exit "$awk_status"
++    fi
++    julia --project=. /tmp/cat_elements_dynamic.jl
++    ;;
++  *)
++    echo "usage: $0 base|new"
++    exit 2
++    ;;
++esac

tasks/format-code-task-000020/tests/test.sh

@@ -0,0 +1,38 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+test/categorical_algebra/CatElements.jl
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000020/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh base && bash /testbed/mimo_test_command.sh new

tasks/format-code-task-000027/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:b342c819994822f62255717bd8ed7917e91965ca9bc7fab36f339edd596bdb2f

tasks/format-code-task-000027/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000027/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000027/tests/test.patch

@@ -0,0 +1,78 @@
+diff --git a/test/types/schema.test.ts b/test/types/schema.test.ts
+index fc0204e2a..fc7ae0fa0 100644
+--- a/test/types/schema.test.ts
++++ b/test/types/schema.test.ts
+@@ -1230,3 +1230,51 @@ async function gh13797() {
+     expectType<IUser>(this); return '';
+   } } });
+ }
++
++function gh13900StringMapOfNumber() {
++  const schema = new Schema({
++    myMap: { type: 'Map', of: Number, required: true }
++  });
++  const Test = model('Test_gh13900_num', schema);
++
++  const doc = new Test({ myMap: { answer: 42 } });
++  expectType<Map<string, number>>(doc.myMap);
++  expectType<number | undefined>(doc.myMap!.get('answer'));
++}
++
++function gh13900StringMapOfString() {
++  const schema = new Schema({
++    myMap: { type: 'Map', of: String, required: true }
++  });
++  const Test = model('Test_gh13900_str', schema);
++
++  const doc = new Test({ myMap: { answer: 'hi' } });
++  expectType<Map<string, string>>(doc.myMap);
++  expectType<string | undefined>(doc.myMap!.get('answer'));
++}
++
++function gh13900StringMapOfMixed() {
++
++  const schema = new Schema({
++    instance: {
++      type: 'Map',
++      of: Schema.Types.Mixed,
++      required: true
++    }
++  });
++  const Test = model('Test_gh13900_mixed', schema);
++
++  const doc = new Test({ instance: { foo: 1 } });
++  expectType<Map<string, any>>(doc.instance);
++}
++
++function gh13900StringMapPreservesConstructorBehavior() {
++
++  const schema = new Schema({
++    myMap: { type: Map, of: Number, required: true }
++  });
++  const Test = model('Test_gh13900_ctor', schema);
++
++  const doc = new Test({ myMap: { x: 1 } });
++  expectType<Map<string, number>>(doc.myMap);
++}
+diff --git a/test_commands.json b/test_commands.json
+new file mode 100644
+index 000000000..0f7dbefd0
+--- /dev/null
++++ b/test_commands.json
+@@ -0,0 +1,3 @@
++{
++  "test_commands": ["npx tsd"]
++}
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,8 @@
++#!/usr/bin/env bash
++
++set -uo pipefail
++cd /testbed
++rc=0
++npx tsd
++rc=$(( rc | $? ))
++exit $rc

tasks/format-code-task-000027/tests/test.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+test/types/schema.test.ts
+test_commands.json
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000027/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000029/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:83c68bed4b1c6ec51847df71e135dd5635b377dfabfcc7f856f03e6203e1621e

tasks/format-code-task-000029/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000029/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000029/tests/test.patch

@@ -0,0 +1,79 @@
+diff --git a/test/model.test.js b/test/model.test.js
+index c2e3f68c1e8..cca70e32fd8 100644
+--- a/test/model.test.js
++++ b/test/model.test.js
+@@ -5891,6 +5891,54 @@ describe('Model', function() {
+     assert.equal(called, 1);
+   });
+ 
++  it('custom statics that overwrite model functions dont get hooks by default', async function() {
++
++    const schema = new Schema({ name: String });
++
++    schema.statics.insertMany = function(docs) {
++      return model.insertMany.apply(this, [docs]);
++    };
++
++    let called = 0;
++    schema.pre('insertMany', function(next) {
++      ++called;
++      next();
++    });
++    const Model = db.model('Test', schema);
++
++    const res = await Model.insertMany([
++      { name: 'foo' },
++      { name: 'boo' }
++    ]);
++
++    assert.ok(res[0].name);
++    assert.ok(res[1].name);
++    assert.equal(called, 1);
++  });
++
++  it('custom statics that overwrite document functions dont get hooks by default', async function() {
++
++    const schema = new Schema({ name: String });
++
++    schema.statics.save = function() {
++      return 'foo';
++    };
++
++    let called = 0;
++    schema.pre('save', function(next) {
++      ++called;
++      next();
++    });
++
++    const Model = db.model('Test', schema);
++
++    const doc = await Model.save();
++
++    assert.ok(doc);
++    assert.equal(doc, 'foo');
++    assert.equal(called, 0);
++  });
++
+   it('error handling middleware passes saved doc (gh-7832)', async function() {
+     const schema = new Schema({ _id: Number });
+ 
+diff --git a/mimo_build_env.tar.gz.b64 b/mimo_build_env.tar.gz.b64
+new file mode 100644
+--- /dev/null
++++ b/mimo_build_env.tar.gz.b64
+@@ -0,0 +1,4 @@
++H4sIAAAAAAAAA+3RUYrCMBSF4Ty7irsBY9KayXKkNgE7mGawqbj8afsyIji+KMPg/72cC7mQC0fv
++x+4YdrE/b9SrmIl3bsnJbS6zdZWtjNk6XyljrfNeiXvZRVfGoTQnEXXKufy29+j9n9I//Zc4lF2b
++U2r6oIfD8/541L+97r/2U//bD1MrMc874b43778NshS/j2HVf10k5fbQyHodL12ZonQp5rFIPfck
++8+Im5RCPeh7157D66/sBAAAAAAAAAAAAAACAd/YNrbE5AwAoAAA=
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,6 @@
++#!/usr/bin/env bash
++
++set -uo pipefail
++cd /testbed || exit 1
++base64 -d mimo_build_env.tar.gz.b64 | tar -xzf - -C /testbed || exit 1
++exec bash -lc 'cd /testbed && bash -eo pipefail /testbed/.build_env/test_command.sh'

tasks/format-code-task-000029/tests/test.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+test/model.test.js
+mimo_build_env.tar.gz.b64
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000029/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000042/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:df6d414cfb63da9b2b2766392a7de94f58a57b4f6d1656766599ceace007f75c

tasks/format-code-task-000042/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000042/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000042/tests/test.patch

@@ -0,0 +1,55 @@
+diff --git a/test/data/dataset/test_dataset.py b/test/data/dataset/test_dataset.py
+index 3d7e6b776..caa299491 100644
+--- a/test/data/dataset/test_dataset.py
++++ b/test/data/dataset/test_dataset.py
+@@ -14,7 +14,7 @@
+ from cartoframes.auth import Credentials
+ from cartoframes.data.clients import SQLClient
+ from cartoframes.utils.geom_utils import setting_value_exception
+-from cartoframes.utils.columns import normalize_name
++from cartoframes.utils.columns import normalize_name, DataframeColumnsInfo
+ from cartoframes.utils.utils import load_geojson
+ from cartoframes.data import StrategiesRegistry
+ from cartoframes.data.dataset.registry.dataframe_dataset import DataFrameDataset, _rows
+@@ -906,3 +906,14 @@ def test_rows_null_geom(self):
+         rows = _rows(df, dataframe_columns_info, with_lnglat)
+ 
+         self.assertEqual(list(rows), [b'|\n', b'|\n'])
++
++    def test_rows_non_ascii(self):
++        attribute = 'áéí'
++        unicode_attribute = u'áéí'
++        encoded_attribute = unicode_attribute.encode('utf-8')
++        encoded_line = encoded_attribute + '\n'.encode()
++
++        df = pd.DataFrame.from_dict({'test': [attribute, unicode_attribute, encoded_attribute, 'xyz']})
++        columns_info = DataframeColumnsInfo(df)
++        rows = _rows(df, columns_info, None)
++        self.assertEqual(list(rows), [encoded_line, encoded_line, encoded_line, b'xyz\n'])
+diff --git a/mimo_build_env.tar.gz.b64 b/mimo_build_env.tar.gz.b64
+new file mode 100644
+--- /dev/null
++++ b/mimo_build_env.tar.gz.b64
+@@ -0,0 +1,11 @@
++H4sIAAAAAAAAA+2UT2/TQBDFc/anGFqkwsFJ7DaphChSxR+1F1qVcEBCqta7k3gbZ3e7u04aUb47
++YyehBVF6aQWI+R3seD1583beJt2i1pU6RzPvdR6LPrE/GLR34ud7+zkb5FmeDwb53n6nn2XD/rAD
++g0dzdIs6ROEBOt7a+Lu6+97/o3Rv8o8Y4rm0s5kwqhvKh+txT/5ZPsy+559lu5T/3jDLO9B/OAt3
++85/nv/2kV2jTK0QoE6mgPQQFqiTZhrcm1B7hdBlLa2C3m+WgAwgZ9RwThfM0LHSUJbi2YHdVkb/q
++0aueqasKrq8h+hobrWNDY6YlhQ6NQiM1BtBjMDaCqDwKtQTnMaCJyUYvnYHTDvT6q2laUN00DcsQ
++cZY6IadiQirpZeOAbIOjgysCSOGjhVAKh9US4qWaQW20QmkVwgILaSvrA1xocyFyEM4pTY9uKby3
++C5igXevsBIy1i9ZW4eV+f+eurX3ACKefRkcn708PR0dArbWRVU29Yom0KXuBMsKzVqzrllCStIDC
++2ykamKMPmqYbHEo91uifJ3jlrL8tebAJ5cXTLzerX5veZ7Vpu9ARnpCNdgpjXSGkgFcrEyNaeyOi
++oP5kLOLEi9h0bGoDLEpNCXq8rDVF/frwbHQC0iMlFLWowg9ZrIbcXHqKBNsLqa7+N9YPzQY/J0Ck
++U9ja9H7nxQzXJj4aHcH6275+sXRsxnZrozSn7GNxEEqaS/KnfzAMwzAMwzAMwzAMwzAMwzAMwzAM
++wzAMwzB/Gd8AiYEWdwAoAAA=
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,6 @@
++#!/usr/bin/env bash
++
++set -uo pipefail
++cd /testbed || exit 1
++base64 -d mimo_build_env.tar.gz.b64 | tar -xzf - -C /testbed || exit 1
++exec bash -lc 'cd /testbed && bash -eo pipefail /testbed/.build_env/test_command.sh'

tasks/format-code-task-000042/tests/test.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+test/data/dataset/test_dataset.py
+mimo_build_env.tar.gz.b64
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000042/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000045/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:fcb5d0910fc4725ca505d3cfb8e05b990140a83fe1c17f0152db8c742982feab

tasks/format-code-task-000045/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com

tasks/format-code-task-000045/environment/setup/setup.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+# Runs once, as root, before the agent starts: Harbor's [environment.healthcheck] unpacks this folder into
+# /var/lib/mimo and runs it. Generated by mimo_harbor; see its README for how each step maps to Xiaomi's harness.
+M=/var/lib/mimo
+[ -f "$M/ready" ] && exit 0
+chmod 700 "$M"
+fail() { echo "setup: $*" >&2; exit 1; }
+# the answer-leak blocklist (mimoagent), applied to /etc/hosts by the agent right after it installs
+write_blocklist() { cp "$M/files/blocklist" "$M/blocklist"; }
+
+# Code: mimoagent environments/datasets/opensource_code.py + base.py, as the explorer runs it (app/runner/domains.py Code).
+CWD=/testbed
+git config --global --add safe.directory "$CWD"
+cd "$CWD" || fail "no working directory $CWD"
+if ! git rev-parse --git-dir >/dev/null 2>&1; then   # a bare source tree: give it a baseline commit
+  git init -q && git add -A && git -c user.name=mimo -c user.email=mimo@localhost commit -q -m baseline --allow-empty
+fi
+BASE=$(git rev-parse HEAD 2>/dev/null)
+[ ${#BASE} -eq 40 ] || fail "could not resolve the base commit in $CWD"
+echo "$BASE" > "$M/base"
+# Images are built with history truncated at the base. If one is not, the fix could be read out of git log,
+# so .git is hidden while the agent works and put back for grading.
+LATER=$(git rev-list --all --not "$BASE" 2>/dev/null | head -n 5 | grep -c . || true)
+rm -f /tmp/fail.log /tmp/pass.log /tmp/patch.diff /tmp/test_patch.diff /tmp/*.log 2>/dev/null; rm -rf /tmp/claude-0 /tmp/claude-* /tmp/testem-* /tmp/puppeteer_dev_chrome_profile-* 2>/dev/null; rm -f /tmp/test_files.json 2>/dev/null; rm -rf /tmp/jest_* /tmp/jest-* /tmp/build /tmp/build_env 2>/dev/null; rm -rf /tmp/pytest-of-root /tmp/pytest-* /tmp/.pytest_cache /tmp/__pycache__ 2>/dev/null; rm -rf /tmp/go-build* /tmp/standards /tmp/wordpress /tmp/zig-* /tmp/testbase /tmp/testbed 2>/dev/null; rm -f /tmp/tmp*.tmp /tmp/*.bak /tmp/expect* /tmp/butwas* 2>/dev/null; rm -rf /root/.cache/go-build 2>/dev/null; rm -rf /tests 2>/dev/null; find /logs -mindepth 1 -delete 2>/dev/null; mkdir -p /logs/agent /logs/verifier /logs/artifacts; chmod 777 /logs/agent /logs/verifier /logs/artifacts; find /var/log -type f -delete 2>/dev/null || true; rm -rf /var/lib/postgresql/*/*/log /var/lib/postgresql/*/*/pg_log 2>/dev/null || true; find /var/lib/mysql /var/lib/mongodb -type f -name '*.log' -delete 2>/dev/null || true; rm -rf /root/.npm/_logs /root/.babel.json /root/.pytest_cache 2>/dev/null || true; find / -maxdepth 4 -xdev -name task_description.md -path '*/.build_env/*' -delete 2>/dev/null || true; true
+git clean -fdx --exclude=node_modules --exclude=bower_components --exclude=.husky --exclude=vendor --exclude=third_party --exclude=_deps --exclude=vcpkg_installed --exclude=.venv --exclude=venv --exclude=.gradle --exclude=target --exclude=.build --exclude=lib --exclude=.bundle --exclude=Manifest.toml --exclude=_build --exclude=.stack-work --exclude=dist-newstyle >/dev/null 2>&1
+find /root/.m2 -type f \( -name '*-SNAPSHOT.jar' -o -name '*-SNAPSHOT-sources.jar' -o -name '*-SNAPSHOT-tests.jar' -o -name '*-SNAPSHOT-test-sources.jar' \) -delete 2>/dev/null; rm -rf /root/.julia/compiled 2>/dev/null; rm -rf /root/.gradle/caches/build-cache-* /root/.gradle/daemon 2>/dev/null; rm -rf /root/.gradle/caches/*/scripts /root/.gradle/caches/jars-* 2>/dev/null; rm -rf /root/.cache/bazel 2>/dev/null; true
+if [ "$LATER" -gt 0 ]; then
+  mv "$CWD/.git" "$M/git-hidden"
+  echo "history not truncated at ${BASE:0:12}: .git hidden while the agent works"
+fi
+write_blocklist
+
+touch "$M/ready"
+echo "setup done"

tasks/format-code-task-000045/tests/test.patch

@@ -0,0 +1,120 @@
+diff --git a/test/BootBot.spec.js b/test/BootBot.spec.js
+index 39f0157..68df63c 100644
+--- a/test/BootBot.spec.js
++++ b/test/BootBot.spec.js
+@@ -116,6 +116,93 @@ describe('BootBot', () => {
+     expect(spy.calledWith(expected)).to.equal(true);
+   });
+ 
++  it('formats a string quick reply with default content_type, title, and payload', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendTextMessage(userId, 'Hello', ['Red']);
++    const body = spy.firstCall.args[0];
++    expect(body.message.quick_replies).to.deep.equal([
++      { content_type: 'text', title: 'Red', payload: 'BOOTBOT_QR_RED' }
++    ]);
++  });
++
++  it('passes through a location quick reply object as-is', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendTextMessage(userId, 'Where are you?', [{ content_type: 'location' }]);
++    const body = spy.firstCall.args[0];
++    expect(body.message.quick_replies).to.deep.equal([
++      { content_type: 'location' }
++    ]);
++  });
++
++  it('preserves all fields of a text quick reply with image_url', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    const reply = {
++      content_type: 'text',
++      title: 'Green',
++      payload: 'DEVELOPER_DEFINED_PAYLOAD_FOR_PICKING_GREEN',
++      image_url: 'http://petersfantastichats.com/img/green.png'
++    };
++    bot.sendTextMessage(userId, 'Pick a color', [reply]);
++    const body = spy.firstCall.args[0];
++    expect(body.message.quick_replies).to.deep.equal([reply]);
++  });
++
++  it('auto-generates payload and content_type for an object with only a title', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendTextMessage(userId, 'Pick', [{ title: 'Red' }]);
++    const body = spy.firstCall.args[0];
++    expect(body.message.quick_replies).to.deep.equal([
++      { content_type: 'text', title: 'Red', payload: 'BOOTBOT_QR_RED' }
++    ]);
++  });
++
++  it('preserves user-provided payload when title is supplied', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendTextMessage(userId, 'Pick', [{ title: 'Red', payload: 'CUSTOM_RED_PAYLOAD' }]);
++    const body = spy.firstCall.args[0];
++    const qr = body.message.quick_replies[0];
++    expect(qr.title).to.equal('Red');
++    expect(qr.payload).to.equal('CUSTOM_RED_PAYLOAD');
++    expect(qr.content_type).to.equal('text');
++  });
++
++  it('formats a mixed array of string and object quick replies correctly', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendTextMessage(userId, 'Pick', [
++      'Red',
++      { content_type: 'location' },
++      { content_type: 'text', title: 'Blue', payload: 'PICK_BLUE', image_url: 'http://x/blue.png' }
++    ]);
++    const body = spy.firstCall.args[0];
++    expect(body.message.quick_replies).to.deep.equal([
++      { content_type: 'text', title: 'Red', payload: 'BOOTBOT_QR_RED' },
++      { content_type: 'location' },
++      { content_type: 'text', title: 'Blue', payload: 'PICK_BLUE', image_url: 'http://x/blue.png' }
++    ]);
++  });
++
++  it('preserves extra fields on a location quick reply', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendTextMessage(userId, 'Where?', [{ content_type: 'location', extra_field: 'KEEP_ME' }]);
++    const body = spy.firstCall.args[0];
++    const qr = body.message.quick_replies[0];
++    expect(qr.content_type).to.equal('location');
++    expect(qr.extra_field).to.equal('KEEP_ME');
++  });
++
++  it('formats quick replies on sendAttachment the same way as sendTextMessage', () => {
++    const spy = sinon.spy(bot, 'sendRequest');
++    bot.sendAttachment(userId, 'image', 'http://x/img.png', [
++      { content_type: 'location' },
++      { title: 'Red' }
++    ]);
++    const body = spy.firstCall.args[0];
++    expect(body.message.quick_replies).to.deep.equal([
++      { content_type: 'location' },
++      { content_type: 'text', title: 'Red', payload: 'BOOTBOT_QR_RED' }
++    ]);
++  });
++
+   it('can send an attachment', () => {
+     const spy = sinon.spy(bot, 'sendRequest');
+     const type = 'image';
+diff --git a/test_commands.json b/test_commands.json
+new file mode 100644
+index 0000000..fdb8120
+--- /dev/null
++++ b/test_commands.json
+@@ -0,0 +1,3 @@
++{
++  "test_commands": ["npx mocha test/BootBot.spec.js"]
++}
+diff --git a/mimo_test_command.sh b/mimo_test_command.sh
+new file mode 100755
+--- /dev/null
++++ b/mimo_test_command.sh
+@@ -0,0 +1,8 @@
++#!/usr/bin/env bash
++
++set -uo pipefail
++cd /testbed
++rc=0
++npx mocha test/BootBot.spec.js
++rc=$(( rc | $? ))
++exit $rc

tasks/format-code-task-000045/tests/test.sh

@@ -0,0 +1,39 @@
+#!/bin/bash
+# Code grading, mimoagent opensource_code.py: reset the files the hidden test patch touches, apply it, run the
+# task's test command; reward 1 when it exits 0. A testbed failure (reset or apply) writes no reward: not scored.
+M=/var/lib/mimo
+CWD=/testbed
+V=/logs/verifier
+mkdir -p "$V"
+BASE=$(cat "$M/base" 2>/dev/null) || { echo "setup never ran (no $M/base)" >&2; exit 1; }
+cd "$CWD" || exit 1
+if [ -d "$M/git-hidden" ]; then rm -rf "$CWD/.git"; mv "$M/git-hidden" "$CWD/.git"; fi
+# the agent's change against the base, so its own commits count too
+git add -A >/dev/null 2>&1 && git -c core.fileMode=false diff --cached "$BASE" > "$V/agent.diff"
+git reset -q
+if ! (
+while IFS= read -r tf; do
+  [ -z "$tf" ] && continue
+  if git cat-file -e $BASE:"$tf" 2>/dev/null; then
+    git checkout $BASE -- "$tf" 2>/dev/null || true
+  else
+    git rm -f --cached "$tf" >/dev/null 2>&1 || true
+    rm -f "$tf"
+  fi
+done <<'EOF_RESET_TEST_FILES'
+test/BootBot.spec.js
+test_commands.json
+mimo_test_command.sh
+EOF_RESET_TEST_FILES
+
+); then echo "could not reset the test files (testbed problem, not scored)" >&2; exit 1; fi
+if ! git apply --verbose /tests/test.patch > "$V/apply.log" 2>&1; then
+  cat "$V/apply.log" >&2
+  echo "the hidden tests could not be applied (testbed problem, not scored)" >&2
+  exit 1
+fi
+timeout 1800 sh -c "$(cat /tests/test_command.sh)" > "$V/test_output.log" 2>&1
+RC=$?
+tail -c 12000 "$V/test_output.log"
+if [ $RC -eq 0 ]; then echo 1 > "$V/reward.txt"; else echo 0 > "$V/reward.txt"; fi
+echo "test command exited $RC"

tasks/format-code-task-000045/tests/test_command.sh

@@ -0,0 +1 @@
+bash /testbed/mimo_test_command.sh

tasks/format-code-task-000089/environment/Dockerfile

@@ -0,0 +1,3 @@
+# The task's image, pinned by digest (the same as [environment].docker_image).
+# Setup runs at start from [environment.healthcheck]; environment/setup/ is its readable copy.
+FROM docker.io/xiaomimimo/mimo-v2.6-rl-oss@sha256:02c1b61699c8ba5562395a4c5ba9f565c5d5a31733ae3fe5603ddb069efea7c9

tasks/format-code-task-000089/environment/setup/files/blocklist

@@ -0,0 +1,33 @@
+# mimo answer-leak blocklist (mimoagent)
+0.0.0.0 github.com
+0.0.0.0 www.github.com
+0.0.0.0 api.github.com
+0.0.0.0 raw.githubusercontent.com
+0.0.0.0 codeload.github.com
+0.0.0.0 objects.githubusercontent.com
+0.0.0.0 gist.github.com
+0.0.0.0 gist.githubusercontent.com
+0.0.0.0 gitlab.com
+0.0.0.0 bitbucket.org
+0.0.0.0 sourceforge.net
+0.0.0.0 git.kernel.org
+0.0.0.0 chromium.googlesource.com
+0.0.0.0 android.googlesource.com
+0.0.0.0 sourcegraph.com
+0.0.0.0 bugs.chromium.org
+0.0.0.0 issues.oss-fuzz.com
+0.0.0.0 oss-fuzz.com
+0.0.0.0 osv.dev
+0.0.0.0 api.osv.dev
+0.0.0.0 bugzilla.mozilla.org
+0.0.0.0 proxy.golang.org
+0.0.0.0 sum.golang.org
+0.0.0.0 pkg.go.dev
+0.0.0.0 google.com
+0.0.0.0 www.google.com
+0.0.0.0 bing.com
+0.0.0.0 www.bing.com
+0.0.0.0 duckduckgo.com
+0.0.0.0 html.duckduckgo.com
+0.0.0.0 search.brave.com
+0.0.0.0 stackoverflow.com