text stringlengths 185 73.3k | repo stringlengths 7 100 | path stringlengths 4 146 | language stringclasses 7
values | hash stringlengths 16 16 | score float64 7 8.5 | stars int64 0 237k |
|---|---|---|---|---|---|---|
"""One map, run wide: the parallelism primitive the tools share (ADR-026).
from .parallel import pmap
results = pmap(render_one, units) # ordered, like map()
Everything Luria does is embarrassingly parallel at some unit — a scheme's
view, a journal's books, a file's scan, a URL's probe — and none of ... | dmarx/luria | luria/parallel.py | .py | 57a48670b48b414c | 7.64 | 18 |
#!/usr/bin/env python3
"""What *this* project's record is made of, rendered from its own config.
luria index # → <docs>/record.md, with every other view
`configuration.md` is the schema: every key Luria accepts, what it defaults
to, what it means. This page is the other half of the question, and t... | dmarx/luria | luria/record_doc.py | .py | 920fb28bc57b97dc | 7.64 | 18 |
"""What a status *means* in one scheme, declared beside the records.
[ADR-003](../record/decisions.d/ADR-003.md) closed the status vocabulary to
five words and put a lint behind it, on the strength of an audit finding that
every surface guarded by an executable check had held and every surface
governed by prose conven... | dmarx/luria | luria/statuses.py | .py | 47c49d9e629d8f4f | 7.64 | 18 |
"""Build a scheme directory with documents of chosen statuses.
The tests that exercise retired-document reporting used to lean on whatever the
corpus happened to contain, which made them silently weaker as the corpus
changed — and impossible to write at all for a project (like this one) whose
every decision is Active.... | dmarx/luria | tests/_scheme.py | .py | 04021b972bb9c296 | 7.14 | 18 |
"""Shared fixtures.
Every test runs against *this* repo's record, because Luria's first consumer is
Luria ([ADR-009](../record/decisions.d/ADR-009.md)) — a
check that passes on a synthetic fixture and fails on a real corpus has told you
nothing. Tests that need a controlled tree build one and repoint the config at
it ... | dmarx/luria | tests/conftest.py | .py | 54dcfd95e24ea714 | 8.14 | 18 |
"""Undecided decisions, aged
([ADR-035](../record/decisions.d/ADR-035.md)).
Age is the whole point of this report, so the clock is injected: `--as-of` (and
the `today` argument these tests pass) keeps it from being a test that fails on
a Tuesday in November.
"""
import datetime as dt
import sys
from pathlib import Pat... | dmarx/luria | tests/test_adr_pending.py | .py | 0b60bd924f28a204 | 8.14 | 18 |
"""The two collection shapes (ADR-028): append for narrative logs, changelog
for release logs. Pure-function tests — `collect` takes text and bodies."""
from luria.collect import collect
APPEND_VIEW = """# Log
Old entry.
<!-- luria-insert-here -->
"""
CHANGELOG_VIEW = """# Changelog
Assembled from fragments.
<!--... | dmarx/luria | tests/test_collect.py | .py | d503d05cb62afc05 | 8.14 | 18 |
"""Merge-allocated schemes: temporary codes, concretization, aliases (ADR-049).
Every test here runs the loop an adopter runs: mint on a branch, cite while
the context is loaded, concretize where merges serialize, and keep the old
name resolving forever. The fixture builds a real record and drives the real
commands — ... | dmarx/luria | tests/test_concretize.py | .py | 2b80abef11d12642 | 8.14 | 18 |
"""The two merge rules, split by what a table is (ADR-047).
A settings table merges per key — setting `docs` must not clear `reports`. A
family table is replaced whole when declared — its entries are named by the
project, and "you get the ones you wrote" is the only reading under which a
family can shrink. Every case ... | dmarx/luria | tests/test_config.py | .py | 99ed8660d077fd7d | 8.14 | 18 |
"""Files are UTF-8 by construction; the console is the platform's.
A Windows terminal at cp1252 could not encode the arrow in `luria init → path`
and could not write the check mark a status report ends on, so a scaffold
became a stack trace on a machine where nothing was wrong (#112). Worse, the
two halves disagreed: ... | dmarx/luria | tests/test_encoding.py | .py | a37a3b9d6061e1a8 | 8.14 | 18 |
"""`luria init` never overwrites, and says something useful when it skips
the one file an agent reads first (ADR-037)."""
from pathlib import Path
from luria import init
def test_existing_files_are_kept_verbatim(tmp_path, capsys):
(tmp_path / "CLAUDE.md").write_text("mine, hands off\n")
written, skipped, kep... | dmarx/luria | tests/test_init.py | .py | c98974e2a1bf5005 | 8.14 | 18 |
"""Safe inline translations tool
- Generates a unified diff (scripts/translation_patch.diff) with inlined TRANSLATIONS replacements
- Writes a temp file scripts/temp_main_inlined.py for syntax checking and smoke-run
- DOES NOT overwrite main.py
"""
import ast
import difflib
import sys
from pathlib import Path
ROOT = P... | woshiliyihang/prompt-copilot-cli | scripts/inline_translations_safe.py | .py | acd825c96440c2dd | 7.57 | 13 |
from copilot.agent import AgentRuntime
def test_filter_think_tags(tmp_path, monkeypatch):
# Create a minimal AgentRuntime-like object
class Dummy:
def __init__(self):
pass
rt = AgentRuntime.__new__(AgentRuntime)
# Attach the method under test
rt._filter_think_tags = AgentRunti... | woshiliyihang/prompt-copilot-cli | tests/test_think_filter.py | .py | 41b1cabc27c5fa0b | 7.07 | 13 |
import sqlite3
import requests
from bs4 import BeautifulSoup
import scrapy
from scrapy.crawler import CrawlerProcess
# Database setup
DB_PATH = "news.db"
def initialize_db():
"""Ensure the news table has the correct schema, including sub_category."""
conn = sqlite3.connect(DB_PATH)
cursor = conn.cursor()
... | 100rabhsah/ai-news-aggregator | crawler/test3.py | .py | 0088dd29ec42c14f | 7.95 | 7 |
import sqlite3
import requests
from bs4 import BeautifulSoup
import scrapy
from scrapy.crawler import CrawlerProcess
from duckduckgo_search import DDGS
# Database setup
DB_PATH = "newsagent.db"
def initialize_db():
"""Ensure the news table has the correct schema, including sub_category."""
conn = sqlite3.conn... | 100rabhsah/ai-news-aggregator | crawler/test4.py | .py | f637a010b1891e6c | 7.95 | 7 |
import subprocess
import time
import json
import os
# Set timeout (None means no timeout)
SCRAPY_TIMEOUT = 120
progress_file = "progress.json"
temp_news_file = "temp_news.json"
news_file = "news.json"
progress = {"progress": 0, "status": "idle"}
def save_progress():
"""Save progress to progress.json"""
with ... | 100rabhsah/ai-news-aggregator | scraper_runner.py | .py | 23a2b20c87204552 | 7.45 | 7 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""MkDocs hook that expands ``@import "relative/path.md"`` directives.
Why this exists
----------------
Quick-start pages under ``docs/getting-started/`` reuse a set of shared
snippet files (``docs/includes/quick-start/*.md``) via a file-tra... | open-edge-platform/edge-system-qualification | docs/hooks/md_import.py | .py | 9bc57d9367c0409a | 7.42 | 6 |
# Copyright (C) 2026 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""
Metric calculation and aggregation utilities for ChatQnA Core benchmarking.
Provides functions for:
- Token counting using Llama tokenizer
- Percentile calculation
- Throughput and latency aggregation
"""
import logging
from typing impo... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/gen/src/chatqna_core/metrics.py | .py | cd9c396c00988d66 | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""Text Generation preparation functions for assets and environment setup."""
import logging
import os
from typing import Any, Dict
import allure
from sysagent.utils.config import ensure_dir_permissions
from sysagent.utils.core import Resul... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/gen/src/text_generation/preparation.py | .py | b44b202eb555dd1f | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""Text Generation utility functions for device management and cleanup."""
import logging
from typing import Dict, Any, List
from sysagent.utils.infrastructure import DockerClient
logger = logging.getLogger(__name__)
def cleanup_stale_con... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/gen/src/text_generation/utils.py | .py | 0c52e5a5e5c546a1 | 7.42 | 6 |
# Copyright (C) 2026 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""
PMU-based GPU Telemetry Collector - Zero Prerequisites
This module collects GPU metrics directly from Linux PMU without requiring:
- intel_gpu_top binary
- Debugfs access (/sys/kernel/debug)
- Special kernel parameters
- Privileged conta... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/vision/src/containers/openvino_benchmark/pmu_gpu_collector.py | .py | 59f6e9b1fad379a5 | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""
Base Video Analytics Pipeline Benchmark.
This module provides the base class for video analytics benchmarks
including VA multi-stage pipelines (decode/detect/track/classify).
"""
import configparser
import logging
import os
import sys
i... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/vision/src/containers/video_analytics/base_benchmark.py | .py | 87e66f8dccaba545 | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""DLStreamer concurrent analysis and container orchestration functions."""
import grp
import json
import logging
import os
from typing import Any, Dict, List
import docker
import numa
from .container import run_dlstreamer_analyzer_contain... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/vision/src/dlstreamer/concurrent.py | .py | 71a54237e1552670 | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""DLStreamer container management utilities."""
import json
import logging
import os
from sysagent.utils.infrastructure import DockerClient
logger = logging.getLogger(__name__)
def run_video_utils_container(
docker_client: DockerCli... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/vision/src/dlstreamer/container.py | .py | 59c59d4471445130 | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""DLStreamer result processing and validation functions."""
import logging
from typing import Any, Dict
from sysagent.utils.core import Metrics, Result, get_metric_name_for_device
logger = logging.getLogger(__name__)
def process_device_... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/vision/src/dlstreamer/results.py | .py | e5073a329a8b63d2 | 7.42 | 6 |
# Copyright (C) 2025 Intel Corporation
# SPDX-License-Identifier: Apache-2.0
"""DLStreamer utility functions for cleanup, device management, and system operations."""
import concurrent.futures
import json
import logging
import os
from typing import Any, Dict, List, Optional
import docker
from sysagent.utils.core imp... | open-edge-platform/edge-system-qualification | src/esq/suites/ai/vision/src/dlstreamer/utils.py | .py | 72ef0ca0cc24324d | 7.42 | 6 |
# -*- coding: utf-8 -*-
"""经济系统逻辑:金钱存储、增减、排行等(基于 XUID,精确到分)"""
from typing import Any, Callable, Dict, List, Optional
class Economy:
"""经济系统:负责 player_economy 表及金钱相关数据逻辑,不包含 UI 与通知。"""
def __init__(self, database_manager, setting_manager, logger=None):
self.db = database_manager
self.setting_... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/Economy.py | .py | 8228cb22f42e7fbb | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""生物名称翻译:当生物名称中包含 ':' 时视为 MC 未提供对应翻译,从 entity_display_name.txt 读取用户配置的显示名。"""
from pathlib import Path
from typing import Optional
class EntityDisplayNameManager:
"""从 entity_display_name.txt 读取/补写生物显示名。仅对名称中含 ':' 的键进行查询。"""
def __init__(self, base_path: Path, logger=None):
s... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/EntityDisplayNameManager.py | .py | 8a469d090e6ba3be | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""击杀生物金钱奖励:独立配置文件 kill_reward.txt,格式 类型ID=金额(如 minecraft:creeper=10)。"""
import threading
from pathlib import Path
from typing import Dict
def normalize_entity_type_id(entity_type: str) -> str:
"""统一为小写,便于配置键一致。"""
s = str(entity_type or "").strip()
if not s:
return ""
... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/KillRewardConfig.py | .py | 78df053639962e62 | 7.48 | 8 |
from pathlib import Path
MAIN_PATH = 'plugins/ARCCore'
class LanguageManager:
language_dict = {} # Class variable shared across instances
def __init__(self, default_language_code):
self.language_code = default_language_code.upper()
if self.language_code not in LanguageManager.language_dict:
... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/LanguageManager.py | .py | 27fabec1bca6f2fd | 7.48 | 8 |
from pathlib import Path
from typing import Callable, Dict, List
MAIN_PATH = 'plugins/ARCCore'
class SettingManager:
setting_dict = {} # Class variable to store all settings
def __init__(self):
self.setting_file_path = Path(MAIN_PATH) / "core_setting.yml"
self._change_listeners: List[Callabl... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/SettingManager.py | .py | c6ea52b986455362 | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""恶性错误写入 plugins/ARCCore/error_log.txt(与核心配置同目录)"""
import threading
import traceback
from contextlib import suppress
from datetime import datetime
from pathlib import Path
from typing import List, Optional
_file_lock = threading.Lock()
def append_arc_error_log(
log_file_path: str,
e... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/arc_error_log.py | .py | 016bb93791cd6ebf | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""维度标识:统一为官方 namespaced ID(如 minecraft:overworld),支持自定义维度。"""
from __future__ import annotations
from typing import Any, Optional
# 历史写法 → 官方规范 ID(仅用于写入时规范化与一次性库迁移)
_VANILLA_CANONICAL = {
"overworld": "minecraft:overworld",
"minecraft:overworld": "minecraft:overworld",
"Overworld"... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/dimension_utils.py | .py | a3e0bf6935bed35d | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""跨服同步配置:模式解析与分项表映射"""
from typing import Dict, Iterable, List, Literal, Optional, Set, Tuple
SyncConsumerMode = Literal["none", "client", "file"]
# 配置键 -> 同步类别
SYNC_CATEGORY_SETTING_KEYS: Dict[str, str] = {
"SYNC_CLIENT_SYNC_PLAYER": "player",
"SYNC_CLIENT_SYNC_ECONOMY": "economy",
... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/sync_config.py | .py | d2a152cc64499207 | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""跨服数据同步协议定义"""
from enum import IntEnum
from typing import Any, Dict, List, Optional
class SyncMessageType(IntEnum):
"""同步消息类型枚举"""
# 认证相关
AUTH_REQUEST = 0x01 # 客户端认证请求
AUTH_RESPONSE = 0x02 # 认证响应
# 数据操作
QUERY_REQUEST = 0x10 # 查询数据请求
QUE... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/sync_protocol.py | .py | c95df00d406f06e7 | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""本地 DB 变更 → 同步中心镜像(上行)辅助逻辑。"""
from __future__ import annotations
import re
from typing import Any, Dict, List, Optional, Sequence, Tuple
# 可同步表的主键(用于写后回读整行再 upsert)
SYNC_TABLE_PRIMARY_KEYS: Dict[str, Tuple[str, ...]] = {
"player_basic_info": ("xuid",),
"player_economy": ("xuid",),
... | ARC-Minecraft/EndstoneMC-ARC-Core-Plugin | src/endstone_arc_core/sync_write.py | .py | 1b2c6e12ffe4a7a2 | 7.48 | 8 |
# -*- coding: utf-8 -*-
"""The record of what a session handed to its workers.
One line per delegation in `<session>/delegations.jsonl`, written after
the answer has already been shown, so the file is a history rather than
a thing the answer waits on. It holds what the delegation was and what
it cost, never the worker... | oteomamo/SALT | salt/agents/ledger.py | .py | deeec2d178e1583c | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""Who decides how one turn's memory is selected.
Every memory switch already travels as a keyword on the call that uses
it rather than being baked into the session, which is what makes a
per-turn decision possible at all: something can vary a switch for one
selection and leave the session it b... | oteomamo/SALT | salt/agents/policy.py | .py | f019712330f0e40f | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""What a model does with its own reasoning, measured rather than assumed.
Some models let a caller ask for the working or ask for it to be left
out, and they do it through the chat template rather than through
sampling. Others always reason and offer no way not to. Others never
reason at all. ... | oteomamo/SALT | salt/agents/thinking.py | .py | be5d517e30e01ba7 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""The record of the turns a session planned out instead of answering.
One line per agent turn in `<session>/agent_trace.jsonl`, written after
the reply is already on screen, so the file is a history rather than
something the reply waits on.
It holds what the round decided and what each piece ... | oteomamo/SALT | salt/agents/trace.py | .py | 13acd28fa6bca0c1 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""Model registry for saltChat.
Each registered chat model lives under ``salt/models/<alias>/``:
config.json loading + generation settings for the model
weights symlink to the resolved snapshot in the user's HF cache
Weights are never copied: ``register_model`` downloads throu... | oteomamo/SALT | salt/chat/registry.py | .py | 189c99ebfd4a0332 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""saltServe: launch a persistent vllm serve process for saltChat.
Resolves a registered model and execs ``vllm serve`` so the server owns the
model and its prefix cache outlives every saltChat run. The server is a
foreground process in its own terminal - saltChat never manages its
lifecycle, i... | oteomamo/SALT | salt/chat/serve.py | .py | 2e5c5d904df196c3 | 7.54 | 11 |
"""
Short-turn predicates: let terse user decisions ("go with option B") past
the junk filter's length gates at conversation ingest.
"""
import re
from salt.engine.sentence_filter import (
JUNK_CONTAINS,
JUNK_PATTERNS,
MIN_CHAR_LENGTH,
MIN_WORD_COUNT,
contains_url,
is_aggressive_junk,
)
from s... | oteomamo/SALT | salt/chat/shortturn.py | .py | 5a3f9149edda71d5 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""What a refused tool call looks like on the wire.
Every refusal carries a code, and every code renders as a fixed opening
phrase, so a client can tell one kind of refusal from another by reading
the front of the message and a person can read the whole of it. The
phrases are the contract: they... | oteomamo/SALT | salt/mcp/errors.py | .py | 048be08ff6ba525b | 7.54 | 11 |
#!/usr/bin/env python3
"""Synchronize SoRoMoX release metadata and optionally create a release tag.
The script updates the tracked sources of release metadata:
- ``pyproject.toml``
- ``CITATION.cff``
- ``docs/citation.md``
- ``docs/development/changelog.md``
- ``uv.lock``
Generated ``*.egg-info`` metadata is deliber... | tud-phi/soromox | bump_version.py | .py | 91f358dbadb3f0ce | 7.6 | 15 |
"""Batched rendering demo for tendon-actuated PCS robots.
This example vmaps `robot.rollout_to` over randomly sampled, constant tendon
tensions to generate multiple trajectories from the same initial condition,
and visualizes them in a grid using Matplotlib, Open3D, and Viser renderers.
"""
import argparse
from pathl... | tud-phi/soromox | examples/simulation/pcs/simulate_batched_tendon_actuated_pcs.py | .py | d5a002650139d8fb | 7.6 | 15 |
#!/usr/bin/env python3
"""
Changelog parser for extracting release notes from CHANGELOG.md.
This script extracts the changelog content for a specific version
to be used in GitHub releases.
"""
import argparse
import re
import sys
from pathlib import Path
def extract_version_changelog(changelog_path: Path, version: ... | tud-phi/soromox | extract_changelog.py | .py | 145c4ebe827fbf03 | 7.6 | 15 |
"""
Minimal PyElastica rod simulation (NO gymnasium, NO custom elastica_env).
What it does:
- Builds a straight Cosserat rod
- Fixes one end (cantilever)
- Adds linear damping
- Optional gravity
- Optional self-contact
- Adds a simple "tendon-like" actuation (4 channels) that bends the rod by applying
distributed to... | tud-phi/soromox | paper_results/secIVa_benchmarking_sequential_cpu/code/pyelastica/simulate_planar_pyelastica.py | .py | 47959b3bf48aeade | 7.6 | 15 |
"""
Minimal PyElastica rod simulation (NO gymnasium, NO custom elastica_env).
What it does:
- Builds a straight Cosserat rod
- Fixes one end (cantilever)
- Adds linear damping
- Optional gravity
- Optional self-contact
- Adds a simple "tendon-like" actuation (4 channels) that bends the rod by applying
distributed to... | tud-phi/soromox | paper_results/secIVa_benchmarking_sequential_cpu/code/pyelastica/simulate_spatial_pyelastica.py | .py | 55c485f464ba3cb9 | 7.6 | 15 |
#!/usr/bin/env python3
"""Plot soft tentacle residual identification results with standard template aesthetics."""
import argparse
import shutil
import sys
from dataclasses import dataclass
from pathlib import Path
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns
from matpl... | tud-phi/soromox | paper_results/secVa_system_identification/code/plot_system_id_residual.py | .py | 4d8c55565b21e5f9 | 7.6 | 15 |
"""Evaluation metrics reported for the Section V.C control experiments."""
from __future__ import annotations
from collections.abc import Sequence
import numpy as np
from scipy.spatial.transform import Rotation
STEADY_STATE_FRACTION = 0.1
REGULATION_SETPOINT_INTERVALS = (
(3.0, 6.0),
(6.0, 9.0),
(9.0, 1... | tud-phi/soromox | paper_results/secVc_model_based_control/code/evaluation_metrics.py | .py | d1692dc26797d7ff | 7.6 | 15 |
"""Private capacity-admission seam used by the API-only facade."""
from __future__ import annotations
import math
from dataclasses import dataclass
from enum import Enum
from threading import Lock
from typing import Protocol, runtime_checkable
@dataclass(frozen=True)
class _CapacitySnapshot:
"""Immutable observ... | HarvardMadSys/RouteWise | llm_routewise/_capacity_controller.py | .py | 25df501913460d6d | 7.45 | 7 |
"""Dependency-free online output-length estimation for the public facade."""
from __future__ import annotations
import math
from dataclasses import dataclass
@dataclass
class _MeanState:
"""Online arithmetic mean without retaining individual observations."""
count: int = 0
total: float = 0.0
def u... | HarvardMadSys/RouteWise | llm_routewise/_output_length.py | .py | b22725b82a134f52 | 7.45 | 7 |
"""Latency beliefs: rolling empirical profiles plus prior/penalty fallbacks.
This unifies what used to be two implementations of "what do we think this
provider's TTFT looks like right now":
* simulator ``ObservedRollingLatencyProfileStrategy`` — rolling profile with
an oracle fallback to the provider's true distri... | HarvardMadSys/RouteWise | llm_routewise/core/beliefs.py | .py | 9a1273373d4c50cd | 7.45 | 7 |
"""Shared RouteWise effective-cost primitives.
This module is intentionally pure: it does not import provider, policy,
simulator, or real-eval types. Harnesses extract request cost and scarcity
signals from their own mutable state, then delegate the common math here.
"""
from __future__ import annotations
import mat... | HarvardMadSys/RouteWise | llm_routewise/core/cost.py | .py | 5d7275c0f3567295 | 7.45 | 7 |
"""Shared RouteWise hedging primitives.
This module is limited to pure checkpoint/probability math and generic backup
selection. Harnesses own provider inventory, capacity mutation, transport, and
profile storage.
"""
from __future__ import annotations
import math
from dataclasses import dataclass, field
from typing... | HarvardMadSys/RouteWise | llm_routewise/core/hedging.py | .py | a9a4f2b734a2b1c6 | 7.45 | 7 |
"""Shared rolling latency-profile estimator.
This module is part of the environment-agnostic algorithm core: the profile
consumes ``(timestamp, ttft_ms)`` samples and answers mean/CDF queries over a
causal moving window. Both the simulator policies and the live real-eval
harness parameterize the same estimator; only t... | HarvardMadSys/RouteWise | llm_routewise/core/latency_profile.py | .py | 3262d4ad03d05fc6 | 7.45 | 7 |
"""Shared RouteWise budget LP solver.
The RouteWise body router solves a small linear program:
minimize objective · weights
subject to cost · weights <= budget
sum(weights) = 1
weights >= 0
With one budget inequality plus the probability-simplex constraints, an
optimum is ... | HarvardMadSys/RouteWise | llm_routewise/core/lp.py | .py | 841b42f499dbdbf2 | 7.45 | 7 |
"""Shared token-pricing arithmetic for cache-aware request costs.
Both the simulator (per-token prices on ``Provider``) and the real-eval
harness (per-million prices on ``ProviderSpec``) bill a request as
input_price * uncached_input + cached_input_price * cached_input
+ output_price * output
with the cached... | HarvardMadSys/RouteWise | llm_routewise/core/pricing.py | .py | e6064a0f8acbf6aa | 7.45 | 7 |
"""The RouteWise routing algorithm, written once against ProviderView.
This module owns the orchestration shared by RouteWise adapters, including the
paper artifact's simulator and real-evaluation harness:
* body selector: effective-cost map -> range budget
``B_alpha = (1 - alpha) * c_min + alpha * c_max`` -> budge... | HarvardMadSys/RouteWise | llm_routewise/core/router.py | .py | 044a14ff24cf8d94 | 7.45 | 7 |
"""Shared RouteWise decision data contracts.
These types are intentionally dependency-free and side-effect-free. They
describe policy outputs, not provider state or execution lifecycle mutation.
"""
from __future__ import annotations
from dataclasses import dataclass, field
from typing import TYPE_CHECKING, Any
if ... | HarvardMadSys/RouteWise | llm_routewise/core/types.py | .py | 3d35fe0ff7dfea7a | 7.45 | 7 |
"""Public exception hierarchy for the RouteWise library facade."""
from __future__ import annotations
class RouteWiseError(Exception):
"""Base class for errors raised by the public RouteWise API."""
class ValidationError(RouteWiseError, ValueError):
"""Raised when a caller supplies an invalid public-API ar... | HarvardMadSys/RouteWise | llm_routewise/errors.py | .py | 26589d3eb5afd3bb | 7.45 | 7 |
"""Tests for the rolling latency profile's error-aware extensions."""
from __future__ import annotations
import pytest
from llm_routewise.core.latency_profile import RollingLatencyProfile
def _profile_with_samples(
samples: list[tuple[float, float]],
window_sec: float = 100.0,
) -> RollingLatencyProfile:
... | HarvardMadSys/RouteWise | tests/unit/core/test_latency_profile.py | .py | a9057d992bf1e1a0 | 7.95 | 7 |
# -*- coding: utf-8 -*-
"""
领克 App 相关脚本的公共基础模块:H5 / App 原生两套阿里云 API 网关签名算法
(build_signature / build_native_signature),以及 env.json 读写辅助函数
(load_env_data / save_env_fields)。
env.json 结构(三个子对象):
{
"user": {"username": "", "password": "", "token": "", "refreshToken": "", "deviceId": "",
"tokenExp... | shovelshit/LynkCoHelper | LynkCoHelper/lynkco_common.py | .py | 2f692776cad36184 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""
领克App 每日任务编排入口(签到 + 分享 + 积分查询 + 结果通知)。
- run_daily_tasks():编排"查积分 -> 签到 -> 分享 -> 再查积分对比",返回结构化结果字典。
- run_and_notify():加载 token -> 执行每日任务 -> 组装 Markdown -> 推送到 Bark。
签到/分享成功后 myEnergy 接口的积分有几秒异步延迟,故查询"之后积分"前会先 sleep。
用法:
python3 lynkco_daily_tasks.py # 执行每日任务(签到+分享)并推送结果
"""
import... | shovelshit/LynkCoHelper | LynkCoHelper/lynkco_daily_tasks.py | .py | 1ebe4865018a8f42 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""
Bark 推送通知工具模块,提供两个通用能力(不含任何业务逻辑,供
lynkco_daily_tasks.py 按需调用):
- build_markdown_report(result):把任务结果字典组装成 Bark Markdown 文案;
- send_bark_notification(...):把一段 Markdown 文案推送到 Bark。
配置方式:环境变量 LYNKCO_BARK_KEY,或 env.json 的 notify.barkKey 字段
(Bark App「我的」页面可查看),未配置时跳过推送并打印提示,不抛异常。
"""
imp... | shovelshit/LynkCoHelper | LynkCoHelper/lynkco_notify.py | .py | 9d0970c6f75da61a | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""
领克App 分享任务脚本。
分享流程横跨"原生签名"和"H5签名"两套认证体系,接口协议细节、已知限制见
docs/分享任务接口说明.md。重要提醒:接口返回 success 不代表真正加分(每日
有次数上限,重复调用不会重复加分),需自行对比 myEnergy 的 point 字段判断。
do_share() 已封装好"优先简化两步法,失败/无 content_id 时可选回退完整三步法"的
逻辑。通过 lynkco_sign.py / lynkco_daily_tasks.py 运行时会自动执行一次分享;
也可直接运行本文件单独触发一次分享。
"""
import js... | shovelshit/LynkCoHelper | LynkCoHelper/lynkco_share.py | .py | c89caf23fec7cbf7 | 7.54 | 11 |
# -*- coding: utf-8 -*-
"""
领克App 每日签到脚本。签名算法详见 lynkco_common.py / docs/AppSecret_逆向分析记录.md。
使用前提:需要有效 token,由 lynkco_login.py 的 load_token() 统一提供
(自动续期或人工获取),本脚本无需关心 token 具体来源。
2026-07 抓包更新:真正"执行签到"的接口路径已从 /up/api/v1/user/sign 变为
/up/api/v1/user/sign/upgrade,且改走原生 SDK 签名体系(build_native_signature,
NATIVE_APP_KEY/SEC... | shovelshit/LynkCoHelper | LynkCoHelper/lynkco_sign.py | .py | e499d95ac46067fc | 7.54 | 11 |
#!/usr/bin/env python3
"""
Command line interface for Web Forager.
This module provides the entry point for the `web-forager` command.
"""
import argparse
import json
import logging
import sys
from collections.abc import Callable
from .duckduckgo_news import duckduckgo_news_search
from .duckduckgo_search import duckd... | CyranoB/web-forager | src/web_forager/cli.py | .py | 16090de2b0675820 | 7.52 | 10 |
#!/usr/bin/env python3
"""
DuckDuckGo News Search MCP Tool
This tool allows searching for recent news using DuckDuckGo through the MCP framework.
It integrates with the ddgs library's news() method to provide time-sorted news results.
"""
import logging
from typing import Any
from ddgs import DDGS
from ddgs.exceptio... | CyranoB/web-forager | src/web_forager/duckduckgo_news.py | .py | 510d246a142ee020 | 7.52 | 10 |
#!/usr/bin/env python3
"""
Web Fetch — URL content retrieval with automatic fallback.
Tries a direct HTTP fetch with trafilatura for content extraction first.
Falls back to the Jina Reader API when direct fetch fails or returns
insufficient content (e.g., JavaScript-rendered pages, bot-blocked sites).
"""
import json... | CyranoB/web-forager | src/web_forager/web_fetch.py | .py | 3d84717ccc962576 | 7.52 | 10 |
from __future__ import annotations
import asyncio
import json
import os
import shutil
from abc import ABC, abstractmethod
from datetime import datetime
from pathlib import Path
from codejoust.core import AgentRun, AgentSpec
class AgentNotAvailable(RuntimeError):
pass
class AgentAdapter(ABC):
"""Base class... | he-yufeng/CodeJoust | src/codejoust/adapters.py | .py | d80e3f1995a05aa5 | 7.48 | 8 |
from __future__ import annotations
from datetime import datetime
from pathlib import Path
from pydantic import BaseModel, Field
class AgentSpec(BaseModel):
"""An agent that can be asked to solve a task."""
name: str
cli: str
model: str | None = None
extra_args: list[str] = Field(default_factory... | he-yufeng/CodeJoust | src/codejoust/core.py | .py | 4fbe0e7bb54fbe73 | 7.48 | 8 |
from __future__ import annotations
import contextlib
import shutil
import subprocess
from pathlib import Path
class GitError(RuntimeError):
pass
def _git(args: list[str], cwd: Path) -> str:
res = subprocess.run(
["git", *args],
cwd=str(cwd),
check=False,
capture_output=True,... | he-yufeng/CodeJoust | src/codejoust/worktree.py | .py | 5e276e55e7adf681 | 7.48 | 8 |
#!/usr/bin/env python3
"""Static check: every call to _get_friday_system_prompt() must decide
provider/vault_control explicitly. Catches a missing-gate bug BEFORE the rare
code path that would have exercised it ever runs.
Why this exists
----------------
2026-08-25: 22 call sites across the codebase built Friday's sys... | FutureSpeakAI/Agent-Friday | scripts/check_gated_prompt_callers.py | .py | 02ef4d7fb906136f | 7.59 | 14 |
#!/usr/bin/env python3
"""Import smoke test - catches module-level breakage before it costs an outage.
Why this exists
---------------
On 2026-08-19 a registration block in services/agent.py was moved ABOVE the dict
it mutates, creating a module-level use-before-definition. The server could not
import. It died ... | FutureSpeakAI/Agent-Friday | scripts/check_imports.py | .py | 6e7365be998cdedc | 7.59 | 14 |
#!/usr/bin/env python3
"""One-shot Google (Gmail + Calendar) connector for Friday.
Friday's server can read Gmail and Calendar read-only, but only after a Google
OAuth token exists at ``~/.friday/google_token.json``. The OAuth consent step
fundamentally requires *you* to approve access in a browser signed into your
Go... | FutureSpeakAI/Agent-Friday | scripts/friday_google_connect.py | .py | 202a4ddee0fdb5e6 | 7.59 | 14 |
#!/usr/bin/env python
"""Run the test suite so that a failure is impossible to miss.
Why this exists (2026-08-19): sessions verify their work with
pytest tests/ -q | tail -5
A shell pipeline exits with the status of its LAST command, so `tail`'s 0
replaces pytest's 1 and a suite with failures in it reports succe... | FutureSpeakAI/Agent-Friday | scripts/run_tests.py | .py | 61035a4f8863f7f0 | 7.09 | 14 |
# Block ROS 2 launch_testing plugins that try to load at import-time
# and fail because pyyaml isn't available in this venv.
#
# The plugins are registered via setuptools entry-points installed
# system-wide by the ROS 2 Humble packages. We must prevent the
# import from even being attempted by monkeypatching importlib... | guilyx/flybots | conftest.py | .py | 05185ab6d784f0ca | 8.02 | 10 |
# Erwin Lejeune - 2026-02-18
"""Discovery of the runnable simulations.
Simulations live at ``flybots/simulations/<category>/<name>/run.py``. Rather
than maintaining a hand-written registry that drifts out of date, this
module walks the package and reads each simulation's metadata from its own
docstring and README.
"""... | guilyx/flybots | src/flybots/cli/catalogue.py | .py | 5898b537ae8f26de | 7.52 | 10 |
# Erwin Lejeune - 2026-02-18
"""Terminal styling helpers.
Deliberately tiny and dependency-free. Colour is disabled automatically
when the output is not a TTY, when ``NO_COLOR`` is set (see
https://no-color.org), or when ``TERM=dumb`` — so piping ``flybots list``
into a file gives clean text.
"""
from __future__ impo... | guilyx/flybots | src/flybots/cli/console.py | .py | 1b76f19b0c3531f2 | 7.52 | 10 |
# Erwin Lejeune - 2026-08-27
"""Controllers that treat the radio network as part of the plant.
Two problems that look similar and are not:
* **Connectivity maintenance** -- the fleet has a task, and the network must
survive it. Connectivity is a *constraint*.
* **Relay coverage** -- the fleet's task *is* the networ... | guilyx/flybots | src/flybots/comms/controllers.py | .py | b8d3f6c0c3a2d938 | 7.52 | 10 |
# Erwin Lejeune - 2026-08-27
"""Graph metrics for a flying network, and their gradients.
The quantity that matters is the **algebraic connectivity** λ₂: the second
smallest eigenvalue of the weighted graph Laplacian. It is strictly
positive exactly while the network is connected, and it degrades smoothly
as links stre... | guilyx/flybots | src/flybots/comms/graph.py | .py | b5ceddab5768b2c2 | 7.52 | 10 |
# Erwin Lejeune - 2026-08-27
"""Link models: how good is the radio between two aircraft.
Most swarm papers use a *disk* model -- connected inside a radius, not
outside -- which is convenient and produces a discontinuous graph. Every
connectivity controller worth having differentiates the graph, so a hard
disk gives a ... | guilyx/flybots | src/flybots/comms/radio.py | .py | 86eefadddc34a49d | 7.52 | 10 |
# Erwin Lejeune - 2026-02-15
"""Attitude P controller.
Maps desired Euler angles ``[phi_des, theta_des, psi_des]`` and a thrust
scalar to desired body rates via proportional gain, then feeds into the
:class:`RateController` to produce torques.
The D-term is intentionally omitted: the RateController already provides
a... | guilyx/flybots | src/flybots/control/attitude_controller.py | .py | dba4ee95b2e55dd6 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-15
"""Composite flight controller that stacks the four control layers.
Supports multiple :class:`ControlMode` values so callers can inject
commands at any abstraction level (rates, attitude, velocity, or position).
"""
from __future__ import annotations
from enum import Enum, auto
import n... | guilyx/flybots | src/flybots/control/flight_controller.py | .py | 6d8a8b79ecc7f505 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-15
"""Position controller — maps desired position to desired velocity.
Uses measured velocity as the derivative term instead of
finite-differencing position error, which avoids 1/dt noise
amplification and setpoint kick.
Output: ``desired_velocity`` (world frame) for the :class:`VelocityCont... | guilyx/flybots | src/flybots/control/position_controller.py | .py | 2b57a99c0952236f | 7.52 | 10 |
# Erwin Lejeune - 2026-02-15
"""Body-rate P controller (innermost loop).
Maps desired body rates ``[p_des, q_des, r_des]`` to torques
``[tau_x, tau_y, tau_z]`` using proportional gain scaled to the
vehicle inertia.
Gains are kept intentionally conservative to avoid the aggressive
angular accelerations that cause osci... | guilyx/flybots | src/flybots/control/rate_controller.py | .py | 699ecf9f20fd940f | 7.52 | 10 |
# Erwin Lejeune - 2026-02-15
"""Velocity controller — maps desired velocity to attitude + thrust.
This is where the conversion from "twist commands" to desired roll/pitch/yaw
and collective thrust happens, mirroring the velocity-control loop in PX4.
Output: ``(desired_euler, thrust)`` ready for the :class:`AttitudeCo... | guilyx/flybots | src/flybots/control/velocity_controller.py | .py | 818486f77c5428a0 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-18
"""Footprint-aware inflation layer for costmaps.
Uses the vehicle :class:`BaseFootprint` bounding radius to determine
the inflation distance, producing a safety margin that exactly matches
the robot's physical extent.
This is particularly useful for swarm envelopes where a convex hull
foo... | guilyx/flybots | src/flybots/costmap/footprint_layer.py | .py | 281aead5e197203b | 7.52 | 10 |
# Erwin Lejeune - 2026-02-21
"""Obstacle inflation layer for safe-distance costmaps.
Applies a distance-based cost decay around each occupied cell, ensuring
the planner keeps a minimum clearance from obstacles.
Supports two decay modes:
- ``"exponential"``: ``cost = exp(-scaling * d)`` (default, original)
- ``"line... | guilyx/flybots | src/flybots/costmap/inflation_layer.py | .py | 4706895c13f8088a | 7.52 | 10 |
# Erwin Lejeune - 2026-02-17
"""Dynamic social costmap layer based on agent velocity.
Increases cost in front of moving dynamic agents proportional to their
speed, modelling the social discomfort / collision risk of passing close
to a fast-moving entity.
Reference: P. Trautman, A. Krause, "Unfreezing the Robot: Navig... | guilyx/flybots | src/flybots/costmap/social_layer.py | .py | b8bb963756f6ecd7 | 7.52 | 10 |
"""Speed-aware additive layer for 2D costmaps."""
from __future__ import annotations
import numpy as np
from numpy.typing import NDArray
class VelocityCostLayer:
"""Increase local cost proportionally to ego speed.
The layer is blended with an existing costmap by adding a bounded penalty.
This keeps occ... | guilyx/flybots | src/flybots/costmap/velocity_layer.py | .py | cffac8dfede752e6 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-17
"""Procedural building generators for urban environments."""
from __future__ import annotations
import numpy as np
from flybots.environment.obstacles import BoxObstacle
from flybots.environment.world import World
def add_city_grid(
world: World,
n_blocks: tuple[int, int] = (3, ... | guilyx/flybots | src/flybots/environment/buildings.py | .py | 296ac26b453baa9d | 7.52 | 10 |
# Erwin Lejeune - 2026-02-15
"""Environment factories and presets.
Provides ready-made environments (city, indoor, open field) with
matched drone scale suggestions. Every simulation should use one of
these factories to ensure consistent, comparable GIFs.
Usage::
world, buildings = default_world()
world, buil... | guilyx/flybots | src/flybots/environment/default_world.py | .py | 4399e872ba61f2cc | 7.52 | 10 |
# Erwin Lejeune - 2026-02-16
"""Complementary filter for attitude estimation from gyro + accelerometer.
Reference: R. Mahony, T. Hamel, J.-M. Pflimlin, "Nonlinear Complementary
Filters on the Special Orthogonal Group," IEEE TAC, 2008.
DOI: 10.1109/TAC.2008.923738
"""
from __future__ import annotations
import numpy a... | guilyx/flybots | src/flybots/estimation/complementary_filter.py | .py | f16f43801d8c5a62 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-16
"""Extended Kalman Filter for nonlinear state estimation.
Reference: S. Thrun, W. Burgard, D. Fox, "Probabilistic Robotics,"
MIT Press, 2005, Chapter 3.3.
"""
from __future__ import annotations
from collections.abc import Callable
import numpy as np
from numpy.typing import NDArray
cl... | guilyx/flybots | src/flybots/estimation/ekf.py | .py | a704325735c5bb18 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-16
"""Particle filter (sequential importance resampling) for state estimation.
Reference: M. S. Arulampalam et al., "A Tutorial on Particle Filters for
Online Nonlinear/Non-Gaussian Bayesian Tracking," IEEE TSP, 2002.
DOI: 10.1109/78.978374
"""
from __future__ import annotations
from collec... | guilyx/flybots | src/flybots/estimation/particle_filter.py | .py | 7fd22404670e97ae | 7.52 | 10 |
# Erwin Lejeune - 2026-02-22
"""Discrete-time process-noise covariances for constant-velocity models.
A Kalman filter's ``Q`` is the covariance accumulated over **one step**,
so it has to scale with ``dt``. Writing ``Q = diag([...])`` with fixed
numbers is the single most common way to detune a filter: at 200 Hz... | guilyx/flybots | src/flybots/estimation/process_noise.py | .py | 7bd22007bf46e5a8 | 7.52 | 10 |
# Erwin Lejeune - 2026-02-16
"""Unscented Kalman Filter using Van der Merwe scaled sigma points.
Reference: E. A. Wan, R. Van Der Merwe, "The Unscented Kalman Filter for
Nonlinear Estimation," AS-SPCC, 2000. DOI: 10.1109/ASSPCC.2000.882463
"""
from __future__ import annotations
from collections.abc import Callable
... | guilyx/flybots | src/flybots/estimation/ukf.py | .py | 44cff7aafe2b7418 | 7.52 | 10 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.