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 |
|---|---|---|---|---|---|---|
"""Installation state tracking for fast module startup.
Tracks fingerprints of installed modules to skip redundant `uv pip install` calls.
When a module's pyproject.toml/requirements.txt hasn't changed, we can skip
the install step entirely, significantly speeding up startup.
"""
from __future__ import annotations
i... | microsoft/amplifier-foundation | amplifier_foundation/modules/install_state.py | .py | a769fba684cdf5c6 | 7.62 | 16 |
"""Path construction utilities for bundle resources."""
from __future__ import annotations
from pathlib import Path
def construct_agent_path(base: Path, name: str) -> Path:
"""Construct path to an agent file.
Looks for agent in agents/ subdirectory with .md extension.
Args:
base: Base director... | microsoft/amplifier-foundation | amplifier_foundation/paths/construction.py | .py | fd263eee46577877 | 7.62 | 16 |
"""URI parsing and path normalization utilities."""
from __future__ import annotations
import os
import platform
import re
from dataclasses import dataclass
from pathlib import Path
from urllib.parse import urlparse
# Precompiled regex for parsing git+https:// URI paths.
# Extracts path and optional ref (branch/tag)... | microsoft/amplifier-foundation | amplifier_foundation/paths/resolution.py | .py | 743b71f52c15bf9e | 7.62 | 16 |
"""
네이버 카페 공개 게시글 수집 (로그인 불필요)
- 공개 카페만. 비공개 카페는 로그인 필요 → 이 스크립트 범위 밖
- 네이버는 안티봇 강함 → curl_cffi(브라우저 TLS) 또는 playwright 사용
- robots.txt·ToS 확인 필수
"""
from curl_cffi import requests
from bs4 import BeautifulSoup
import json, sys, time, re
HEADERS = {
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
... | dmae97/omk-crawling | examples/naver_cafe_public.py | .py | 4f8bead48ce45262 | 7.42 | 6 |
"""Detection heuristics — what's blocking us, what do we need?"""
from __future__ import annotations
import importlib.util
from dataclasses import dataclass
from enum import Flag, auto
from functools import lru_cache
from urllib.robotparser import RobotFileParser
from omk_crawl.result import CrawlStatus
class Bloc... | dmae97/omk-crawling | omk_crawl/detect.py | .py | cddd6c8a1b89b46f | 7.42 | 6 |
"""Pipeline — compose fetch → extract → convert steps."""
from __future__ import annotations
from collections.abc import Callable
from dataclasses import dataclass, field
from typing import Any
from omk_crawl.result import CrawlResult, CrawlStatus
from omk_crawl.router import SmartRouter
@dataclass
class Pipeline:... | dmae97/omk-crawling | omk_crawl/pipeline.py | .py | cea56acc8a92e2e5 | 7.42 | 6 |
"""Unified crawl result across all tools."""
from __future__ import annotations
import time
from dataclasses import dataclass, field
from enum import Enum
from typing import Any
class CrawlStatus(Enum):
"""Outcome of a crawl attempt."""
OK = "ok"
BLOCKED = "blocked" # 403 / WAF / anti-bot
TLS_BLOC... | dmae97/omk-crawling | omk_crawl/result.py | .py | d602b704f2932e73 | 7.42 | 6 |
"""Detection-aware routing — pick the next tool from *what* is blocking us.
The default escalation chain is a fixed ladder (lightest → heaviest). Once we
detect the blocking mechanism, we can reorder the remaining rungs so the tool
most likely to succeed is tried next — instead of climbing one rung at a time.
That is ... | dmae97/omk-crawling | omk_crawl/routing.py | .py | 85e9a91428712ed9 | 7.42 | 6 |
"""One-time GitHub star nudge shown after a successful interactive run.
Rules this module never breaks:
* stdout is sacred — every byte goes to **stderr**, so `omk-crawl url > out.md`
and `omk-crawl url --json | jq` stay clean.
* Only ever speaks on a real TTY (stdin *and* stderr), never in CI, never under
pytest... | dmae97/omk-crawling | omk_crawl/star.py | .py | ceedce3848f876d6 | 7.42 | 6 |
"""Tool registry — all adapters in one place."""
from __future__ import annotations
from omk_crawl.tools.apk_tool import ApkTool
from omk_crawl.tools.appstore_tool import AppStoreTool
from omk_crawl.tools.autoscraper_tool import AutoscraperTool
from omk_crawl.tools.baemin_tool import BaeminTool
from omk_crawl.tools.b... | dmae97/omk-crawling | omk_crawl/tools/__init__.py | .py | addec7c9ec3e5859 | 7.42 | 6 |
"""APK static analysis adapter — package:// or path to .apk."""
from __future__ import annotations
import json
import time
from pathlib import Path
from typing import Any
from urllib.parse import unquote, urlparse
from omk_crawl.mobile.apk_analyze import analyze_apk
from omk_crawl.result import CrawlResult, CrawlSta... | dmae97/omk-crawling | omk_crawl/tools/apk_tool.py | .py | 912a4cfcfc13bee7 | 7.42 | 6 |
"""Base tool adapter interface."""
from __future__ import annotations
import abc
from typing import Any
from omk_crawl.detect import tool_available
from omk_crawl.result import CrawlResult, CrawlStatus
# Common kwargs every adapter understands *by name*. Each adapter declares which
# it actually implements via `cap... | dmae97/omk-crawling | omk_crawl/tools/base.py | .py | f5244bebfd3331a1 | 7.42 | 6 |
"""InsaneSearch — hyper-aggressive single-URL unblocker.
First-line breaker when everything else fails. Bypasses:
- TLS/JA3 fingerprinting (impersonate rotation × 8 profiles)
- Cloudflare / Turnstile (real browser fallback)
- Akamai / Datadome / Imperva (header trickery)
- Rate limiting (token bucket + jitter)... | dmae97/omk-crawling | omk_crawl/tools/insane_search_tool.py | .py | b420a94fafafc0a7 | 7.42 | 6 |
"""markitdown adapter — file → Markdown conversion."""
from __future__ import annotations
from typing import Any
from omk_crawl.result import CrawlResult, CrawlStatus, _timer
from omk_crawl.tools.base import BaseTool
class MarkitdownTool(BaseTool):
name = "markitdown"
pip_package = "markitdown[all]"
la... | dmae97/omk-crawling | omk_crawl/tools/markitdown_tool.py | .py | 85510eccd8683848 | 7.42 | 6 |
#!/usr/bin/env python3
from __future__ import annotations
import argparse
import re
from datasets import load_dataset
from transformers import AutoModelForCausalLM, AutoTokenizer
# The canonical example deliberately uses Transformers generation, not vLLM.
# TRL 0.20 probes any system-wide vLLM installation while imp... | rwkv-rs/hf-adapter | examples/finetune/grpo_lora.py | .py | d2bc502a265a8e0f | 7.45 | 7 |
#!/usr/bin/env python3
# coding=utf-8
"""Dependency-free, traversal-safe file helpers used by the converter.
``OBSOLETE_08_ADAPTER_FILES`` is only a removal list. It lets an in-place
conversion clean files emitted by the 0.8 performance-oriented layout; none of
those modules is copied, imported, or shipped by the 0.9 ... | rwkv-rs/hf-adapter | rwkv7_hf/adapter_manifest.py | .py | c3a3370ae909804a | 7.45 | 7 |
# coding=utf-8
"""Configuration for the readable, pure-PyTorch RWKV-7 HF reference model."""
from __future__ import annotations
from transformers import PretrainedConfig
class RWKV7Config(PretrainedConfig):
"""Describe an RWKV-7 checkpoint without selecting a hardware backend.
The names intentionally match ... | rwkv-rs/hf-adapter | rwkv7_hf/configuration_rwkv7.py | .py | 6b1b07047312e1fb | 7.45 | 7 |
# coding=utf-8
"""Small, hardware-neutral PyTorch operator boundary for RWKV-7."""
from __future__ import annotations
import torch
def rwkv7_recurrent(
receptance: torch.Tensor,
decay: torch.Tensor,
key: torch.Tensor,
value: torch.Tensor,
a: torch.Tensor,
b: torch.Tensor,
initial_state: t... | rwkv-rs/hf-adapter | rwkv7_hf/ops_rwkv7.py | .py | 4045c41a7b64d372 | 7.45 | 7 |
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Universal MCP Client - Bundle with any skill that needs MCP access.
Supports both HTTP and stdio transports for connecting to MCP servers.
Usage:
# List available tools from an HTTP MCP server
python mcp-client.py list --url http://localhost:8080
# List ... | Ub207/vault-sync | .qwen/skills/browsing-with-playwright/scripts/mcp-client.py | .py | 03bc2c7137b4e2c1 | 7.48 | 8 |
# Copyright Reexpress AI, Inc. All rights reserved.
import json
import os
import numpy as np
import argparse
import time
from pathlib import Path
import codecs
import torch
from collections import namedtuple
from pydantic import BaseModel
import asyncio
def save_by_appending_json_lines(filename_with_path, json_lis... | ReexpressAI/reexpress_mcp_server | code/data_processing/code/analysis/data_utils.py | .py | 2d1de7030bf280dd | 7.48 | 8 |
# Copyright Reexpress AI, Inc. All rights reserved.
# Management of tool-call counts. This implements hard (requiring a server restart) and soft (user resettable) limits
# as a guard against runaway calling of the Reexpress tool and unnecessary sequential calls before adapting the model.
from pathlib import Path
impo... | ReexpressAI/reexpress_mcp_server | code/reexpress/mcp_utils_tool_limits_manager.py | .py | a15ff6c2c636fa93 | 7.48 | 8 |
"""
Reusable FastAPI dependency factories for DB list-endpoint query parameters.
The book filter fields are defined once in BOOK_FILTER_FIELDS. The
book_filters() factory builds a dependency class from them, optionally omitting
fields that are an endpoint's scope rather than a filter (e.g. the plan
endpoint omits "pla... | LibexHQ/Libex | app/api/routes/db/filters.py | .py | 37b000be6b1f3fd3 | 7.5 | 9 |
"""
Shared filter query-parameter dependency for live (Audible-backed) list endpoints.
Exposes only the filters that filter_dicts (app/services/filtering.py) actually
applies to an in-memory book list — so what shows in the OpenAPI docs is exactly
what works. Heavy free-text filters live on /db/book instead, which has... | LibexHQ/Libex | app/api/routes/filter_params.py | .py | 1a5ee19c0574f58b | 7.5 | 9 |
"""
Narrators route schemas.
"""
# Third party
from pydantic import BaseModel
class AudioSampleResponse(BaseModel):
url: str
title: str | None = None
genre: str | None = None
source: str | None = None
class NarratorProfileResponse(BaseModel):
name: str
description: str | None = None
ima... | LibexHQ/Libex | app/api/routes/narrators/schemas.py | .py | 7dee4120873b34ae | 7.5 | 9 |
"""
Core configuration for Libex.
Settings are loaded from environment variables with sensible defaults.
"""
# Standard library
from functools import lru_cache
# Third party
from pydantic_settings import BaseSettings, SettingsConfigDict
class Settings(BaseSettings):
model_config = SettingsConfigDict(
en... | LibexHQ/Libex | app/core/config.py | .py | 38fa89c476c2afcc | 7.5 | 9 |
"""
Custom exceptions for Libex.
"""
class LibexException(Exception):
"""Base exception for all Libex errors."""
def __init__(self, message: str, status_code: int = 500):
self.message = message
self.status_code = status_code
super().__init__(self.message)
class NotFoundException(Libe... | LibexHQ/Libex | app/core/exceptions.py | .py | aa0d3b87700dcdf9 | 7.5 | 9 |
# Standard library
import atexit
import datetime
import logging
import logging.handlers
import os
import queue
import sys
import time
# Local
from app.core.config import get_settings
# Standard library - optional, Unix only. Grouped with the other conditional
# imports rather than above, so the unconditional imports ... | LibexHQ/Libex | app/core/logging.py | .py | facb1c492a8c8ca2 | 7.5 | 9 |
"""
Single source of truth for the libexdb.com migration notice.
`build_migration_notice` is the ONLY predicate for whether the notice is on —
it requires the flag, the new host, and both dates to be present, stripped,
and parseable, with sunset not before announced. Every call site (the
middleware that stamps headers... | LibexHQ/Libex | app/core/migration_notice.py | .py | 3b1424a8f1e86572 | 7.5 | 9 |
"""
Shared utility functions.
"""
# Standard library
import re
from datetime import datetime, time, timedelta, timezone
def strip_html(text: str | None) -> str | None:
"""
Strips HTML tags and cleans up Audible text artifacts.
Handles: HTML tags, escaped quotes, literal \\r\\n sequences,
leading/trai... | LibexHQ/Libex | app/core/utils.py | .py | 10f7fa9e42080be7 | 7.5 | 9 |
"""
Background completion of author-books walks that ran out of time.
A live author-books request is bounded by what a caller will wait for, and
the fronting proxy gives up at 30 seconds regardless. A prolific author's
walk can exceed that, and when it does the request has a partial ASIN list
and a choice: hand it ove... | LibexHQ/Libex | app/services/audible/authors/completion.py | .py | f56d42daea1dffad | 7.5 | 9 |
"""Behavioral tests for the Codex-manifest drift PostToolUse hook.
The hook (``check-codex-manifest-drift.sh``) is a shift-left guard for the CI
``codex-manifest-drift`` gate. It must:
- fire for ANY plugin's ``*/.codex-plugin/plugin.json`` or
``*/.claude-plugin/plugin.json`` edit (not just loom-code),
- derive the... | kouko/monkey-skills | .claude/hooks/test_check_codex_manifest_drift.py | .py | 0fb460d7d43d6ad8 | 7.92 | 6 |
"""Behavioral tests for the memory-store integrity PostToolUse hook.
The hook (``check-memory-store-integrity.sh``) is a shift-left guard for the
CI step that runs ``scripts/check_loom_memory_integrity.py``. It must:
- fire for any edit under ``docs/loom/memory/``, including the store's own
``README.md`` (deleting ... | kouko/monkey-skills | .claude/hooks/test_check_memory_store_integrity.py | .py | 73f8948012b443de | 7.92 | 6 |
"""Tests for the ascii-graph-toolkit SessionStart trigger-card hook.
Task 1 (this file, first test only): pins the hook's emitted JSON shape
and content. Mirrors loom-design/scripts/pipeline/test_family_relay.py's style
(mechanical marker-grep + subprocess execution over the real hook
script, not a mock).
test_descri... | kouko/monkey-skills | ascii-graph-toolkit/scripts/test_trigger_card.py | .py | 10be401d27f9e73b | 7.92 | 6 |
"""align.py — alignment oracle CLI for ASCII/Unicode diagrams.
Verification-class scaffolding: the model draws the diagram, this measures
display-width drift and reports it; the model fixes. It does NOT lay out and
does NOT edit the diagram.
Wires the three drift checks (vertical seam, table equal-width, kink +
arrow... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/align.py | .py | b5eaa273bc4c70b5 | 7.42 | 6 |
"""checks_seam.py — vertical-seam connect check for ASCII/Unicode diagrams.
Verification-class scaffolding: the model draws, this measures and reports
drift, the model fixes. It does NOT lay out and does NOT edit the diagram.
Check: every box vertical (│ ┃ ║ and other line styles) must connect to a
structural glyph (... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/checks_seam.py | .py | b616e35b27943108 | 7.42 | 6 |
"""Table-block equal-width check.
Detect table blocks and flag any row whose display_width differs from
the block's reference width. This catches CJK tables sized by character
count instead of terminal-cell width.
A table block is a maximal run of consecutive lines that share the same
LEFT frame column -- the display... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/checks_table.py | .py | 4d7ef0c2d707d770 | 7.42 | 6 |
"""Linear flow-diagram generator.
render_flow stacks labelled boxes vertically on a common trunk column,
joined by a centered down-arrow:
┌──────────┐
│ 収到訂單 │
└──────────┘
│
▼
┌──────────┐
│ 驗證ユーザー │
└──────────┘
...
All boxes share one interior width (= max labe... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/gen_flow.py | .py | 53953e140c2fdb6d | 7.42 | 6 |
"""Sequence-diagram generator (participants + message rows).
render_seq lays participant boxes side by side across the top, each with a
lifeline stub `┬` centered under its box and vertical lifelines `│` running
below, then renders each message as a label row + a directional arrow row:
┌──────┐ ┌──────────────┐... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/gen_seq.py | .py | 985575b81ccf26f6 | 7.42 | 6 |
"""CJK-aligned ASCII/Unicode table generator.
Column widths and cell padding are computed with display_width (terminal
cells), not str len, so CJK/JP wide-character cells align in a monospace
terminal. Unicode box-drawing characters are used by default; ascii_only
falls back to +, -, | so the output survives ASCII-onl... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/gen_table.py | .py | 9d0616cfc3e1033d | 7.42 | 6 |
"""Classic Unicode tree / hierarchy renderer.
A node is a dict {"label": str, "children": [node, ...]} (children
optional / empty). The root label sits on line 1; each descendant is
prefixed with branch glyphs that precede the label, so CJK labels do
not affect the branch columns:
訂單系統
├─ 訂單服務
└─ 庫存サービス
... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/gen_tree.py | .py | 179b4a8638fd491f | 7.42 | 6 |
"""Dispatch CLI for the four ASCII/Unicode diagram generators.
Reads a JSON payload from stdin describing one shape's input, routes to
the matching generator, prints the rendered diagram, returns 0. Shapes:
table {"headers": [...], "rows": [[...]], "ascii_only": false}
flow {"steps": [...]}
tree {"no... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/generate.py | .py | 0abdf41db583d00e | 7.42 | 6 |
"""Shared display-width primitive for terminal-cell measurement.
Width policy:
- CJK Wide / Fullwidth characters -> 2 cells
- Ambiguous-width characters -> 1 cell
- Box-drawing characters (U+2500..257F) -> 1 cell
- Control / zero-width characters -> 0 cells
- Everything else (narrow AS... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/scripts/width.py | .py | 1624430ac17b300a | 7.42 | 6 |
"""Contract tests for the table-block equal-width check.
A "table block" is a maximal run of consecutive lines that each start
and end with a box vertical (│ as the first and last non-space glyph),
optionally bracketed by ┌─┐ / ├─┤ / └─┘ border lines. find_issues flags
any line in a block whose display_width differs f... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_checks_table.py | .py | f709646f560b0eb4 | 7.92 | 6 |
"""End-to-end dogfood: oracle + generators wired together on mixed 中/日 input.
These are integration smoke tests for the whole ascii-graph loop:
1. A real, display-width-aligned branching flowchart (mixed Traditional
Chinese / Japanese) passes the oracle end-to-end with ZERO drift —
the diagram below was c... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_e2e.py | .py | b0d8e11f0a23bce1 | 7.92 | 6 |
"""Contract tests for the layered-architecture diagram generator.
render_arch stacks one INDEPENDENT box per layer, all sharing one outer
interior width. Correctness is verified by display_width, not len: every
rendered line MUST occupy the same number of terminal cells, and the two
layer boxes MUST share one outer wi... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_gen_arch.py | .py | 132ade78ab54e75f | 7.92 | 6 |
"""Contract tests for the CJK-aligned horizontal bar chart generator.
Label-column alignment is verified by display_width, not len: every
row's label segment MUST occupy the same number of terminal cells so
that CJK/JP/EN-mixed labels line up before the bar starts. Bar lengths
are verified to scale proportionally to t... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_gen_bar.py | .py | 4f00d061e925d673 | 7.92 | 6 |
"""Contract tests for the linear-flow generator (render_flow).
render_flow stacks each step as a box (┌─┐ │ label │ └─┘) on a common
trunk column, joined by a centered down-arrow. The two invariants that
make the diagram look right on a terminal are:
(a) every box-border line shares the same display_width -- so the... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_gen_flow.py | .py | 60ccaeb471d8b185 | 7.92 | 6 |
"""Contract tests for the sequence-diagram generator (participant skeleton).
render_seq lays participant boxes side by side across the top, each with a
lifeline stub `┬` centered under its box and vertical lifelines `│` running
below. Task 1 renders ONLY the participants-only skeleton (messages are
accepted but not ye... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_gen_seq.py | .py | 557d27aa2f369795 | 7.92 | 6 |
"""Contract tests for the CJK-aligned table generator.
Alignment correctness is verified by display_width, not len: every
rendered line MUST occupy the same number of terminal cells so that
CJK/JP/EN-mixed cells line up in a monospace terminal.
"""
import pathlib
import sys
sys.path.insert(0, str(pathlib.Path(__file... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_gen_table.py | .py | a9584d2efffac869 | 7.92 | 6 |
"""Contract tests for the Unicode tree/hierarchy generator.
Branch glyphs (├─ └─ │) precede the labels, so the columns at which
those glyphs appear are constant regardless of CJK label width — that
column-stability is the observable proof the tree was rendered with a
correct per-ancestor continuation prefix.
"""
impo... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_gen_tree.py | .py | dc03424e0a3b386d | 7.92 | 6 |
"""Contract tests for the generate.py dispatch CLI.
render(shape, payload) is the testable seam (no subprocess); it routes
to the matching generator and returns the rendered diagram. Each shape
is exercised with a CJK/JP fixture so we prove the dispatch preserves
the underlying generator's CJK behavior, not just that ... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_generate_cli.py | .py | ae82cad2b7851f53 | 7.92 | 6 |
"""Contract tests for the shared display-width primitive.
Width policy (from the brief): CJK Wide/Fullwidth = 2, Ambiguous = 1,
box-drawing = 1, control/zero-width = 0. Expectations verified against
the wcwidth package before being asserted here.
"""
import pathlib
import sys
sys.path.insert(0, str(pathlib.Path(__fi... | kouko/monkey-skills | ascii-graph-toolkit/skills/ascii-graph/tests/test_width.py | .py | bd91705a445fd8c6 | 7.92 | 6 |
#!/usr/bin/env python3
"""Lint copywriting-toolkit voice anchor library.
Checks each `standards/anchor-{slug}.md` file against the v2 schema defined in
`standards/voice-anchor-meta.md §Schema (v2 — single active schema)`, plus the
machine-checkable portions of the 4-condition anchor selection rubric.
Moved from Pass ... | kouko/monkey-skills | copywriting-toolkit/scripts/lint-anchor-library.py | .py | 8dc9cab6ca1fac8c | 7.42 | 6 |
"""Pins for copywriting-audit-stage's variant re-gating aggregation
semantics (contract/craft convergence vocabulary, plan Task 6).
Window discipline (docs/loom/memory/grep-tests-scope-to-measured-neighborhood.md):
presence assertions are scoped to the "Variant re-gating loop — formal
contract" section (unique anchor ... | kouko/monkey-skills | copywriting-toolkit/scripts/test_audit_gate_semantics.py | .py | ba55aefa99e04c5b | 7.92 | 6 |
"""Pins for check_anchor_copies.py — verifies the two INLINE-duplicated Tier 1
standards files (copywriting-toolkit/CLAUDE.md §INLINE-duplicate clarification)
stay byte-identical across their skill copies.
No registered REQ-ids in this dispatch — @req tags intentionally omitted.
"""
import importlib.util
import sys
f... | kouko/monkey-skills | copywriting-toolkit/scripts/test_check_anchor_copies.py | .py | aed162ce59088fc3 | 7.92 | 6 |
"""Pins for copywriting-toolkit/CLAUDE.md §Envelope Validation (Task 9).
Task 8 shipped `scripts/validate_envelope.py` as the file-borne enforcement
medium for the handoff envelope; this task wires that contract into
CLAUDE.md so the prose that used to only WARN about counters/immutability
now POINTS at the mechanical... | kouko/monkey-skills | copywriting-toolkit/scripts/test_claude_md_validation_wiring.py | .py | e07cf38f792330bb | 7.92 | 6 |
"""Pins for the canonical convergence vocabulary in copywriting-toolkit/CLAUDE.md.
Window discipline (docs/loom/memory/grep-tests-scope-to-measured-neighborhood.md):
every presence assertion is scoped to the vocabulary section extracted by its
unique heading anchor — never whole-file — so generic terms elsewhere in
CL... | kouko/monkey-skills | copywriting-toolkit/scripts/test_convergence_vocabulary.py | .py | 4fb1dc2de572c1de | 7.92 | 6 |
"""Pins for the copywriter-evaluator contract (contract/craft classes +
prior_findings_check round-2 duty).
Window discipline (docs/loom/memory/grep-tests-scope-to-measured-neighborhood.md):
presence assertions are scoped to the evaluator section extracted by its unique
heading anchor — never whole-file — so generic t... | kouko/monkey-skills | copywriting-toolkit/scripts/test_evaluator_contract.py | .py | 60effd95b1e82402 | 7.92 | 6 |
"""Pins for copywriting-form-check-stage's 8b verdict rule (contract-class
aggregation, pointer to CLAUDE.md — no local restatement) and the
form-appropriate-gate rubric's dimension-row class annotations
(contract = objective form-constraint; craft = qualitative, recorded).
Window discipline (docs/loom/memory/grep-tes... | kouko/monkey-skills | copywriting-toolkit/scripts/test_form_check_gate_semantics.py | .py | 01e0ce473987c20e | 7.92 | 6 |
"""Pin: copywriting-intake's Q8 grill BLOCKED halt must be a verifiable
mechanical condition (named still-empty required fields), not a judgment
call ("user cannot decide").
WHY: docs/loom/memory/prose-only-enforcement-dies-on-weak-executors.md —
weak executors preserve vocabulary but drop prose-only enforcement dutie... | kouko/monkey-skills | copywriting-toolkit/scripts/test_intake_blocked_verifiable.py | .py | d2cbbda3b5ebff9d | 7.92 | 6 |
"""Guards the post-draft overlay-mode output-envelope example in
copywriting-neta-injection/SKILL.md against passthrough-field drift.
The canonical envelope shape (copywriting-toolkit/CLAUDE.md §Handoff
Envelope) carries `express_mode_used`, `audit_trail`, and `retries` as
immutable passthrough fields every mid-pipeli... | kouko/monkey-skills | copywriting-toolkit/scripts/test_neta_overlay_example.py | .py | 011d6194fe6df310 | 7.92 | 6 |
"""Pins for copywriting-neta-injection/rubrics/neta-safety-gate.md's
Verdict Rules section: contract-class-only aggregation (mirrors the
sibling fix already landed on voice-consistency-gate.md, commits
50caf36a + e32498db) and copywriting-neta-injection/SKILL.md's verdict
enum declaration.
Window discipline (docs/loom... | kouko/monkey-skills | copywriting-toolkit/scripts/test_neta_rubric_semantics.py | .py | 0d68862451789b6b | 7.92 | 6 |
"""Pins for using-copywriting-toolkit/SKILL.md router validator wiring (Task 10).
Task 9 landed CLAUDE.md's `## Envelope Validation` section (the mechanics:
work-file convention, exit codes, `--prev` duty). This task wires the
ACTING-MOMENT imperative into the router's own SKILL.md — the routing loop
must instruct ser... | kouko/monkey-skills | copywriting-toolkit/scripts/test_router_validator_wiring.py | .py | 8ebdd9e28c71fdb4 | 7.92 | 6 |
"""Pins for copywriting-voice-quadrant-stage's Gate passage: contract-class
semantics applied to declared-voice-target (quadrant) mismatches, with
positioning nuance carried as craft (recorded, never gates).
Window discipline (docs/loom/memory/grep-tests-scope-to-measured-neighborhood.md):
presence assertions are scop... | kouko/monkey-skills | copywriting-toolkit/scripts/test_voice_quadrant_gate_semantics.py | .py | bd484bb269256cb5 | 7.92 | 6 |
"""Pins for copywriting-voice-tone-stage/rubrics/voice-consistency-gate.md's
Verdict Rules section: contract-class-only aggregation (mirrors the sibling
fix already landed in the SKILL.md gate passage, commit 0ea45ea1).
Window discipline (docs/loom/memory/grep-tests-scope-to-measured-neighborhood.md):
presence asserti... | kouko/monkey-skills | copywriting-toolkit/scripts/test_voice_rubric_semantics.py | .py | 9a54bbed0a54122e | 7.92 | 6 |
"""Tier definitions and OnIt agent configuration for benchmarks.
A *tier* is just a preset of (sample limit, concurrency). The benchmark task
code is tier-agnostic; the tier is applied at run time by ``run.py``.
The eval target (model/host) is resolved from environment variables so the same
task code can run against ... | sibyl-oracles/onit | benchmarks/config.py | .py | cc5f236b593f304f | 7.48 | 8 |
"""Inspect AI model provider that drives the real OnIt agent.
Registers an ``onit`` provider so benchmarks can use ``model="onit/<label>"``.
Each :meth:`OnItAPI.generate` call runs one task end-to-end through
:meth:`OnIt.process_task`, exercising OnIt's real prompt engineering, MCP tool
registry, and tool loop — not j... | sibyl-oracles/onit | benchmarks/onit_provider.py | .py | 420cf5ce9ee8bd68 | 7.48 | 8 |
"""Aggregate Inspect ``.eval`` logs into a markdown + JSON summary.
Reads every eval log under a directory, extracts the headline metric per
benchmark, writes ``summary.json`` and ``summary.md`` alongside them, and
(optionally) diffs against a committed baseline to flag regressions. The CI
smoke gate uses ``--baseline... | sibyl-oracles/onit | benchmarks/report.py | .py | 4fac8376afce371c | 7.48 | 8 |
"""CLI entry point for the OnIt benchmark suite.
Thin wrapper over ``inspect_ai.eval`` that:
* registers the ``onit`` model provider,
* resolves the eval target (host/model) from the environment,
* applies a tier preset (sample limit + concurrency),
* and runs one or more benchmark tasks.
Examples:
... | sibyl-oracles/onit | benchmarks/run.py | .py | 4dca08d021a79076 | 7.48 | 8 |
"""SWE-bench runner for the OnIt agent.
SWE-bench is a *repo-editing agent* benchmark, not a code-generation one, so it
does not fit the final-answer provider path used by the other coding tasks. This
runner integrates it the way OnIt actually works:
1. **Edit (OnIt).** For each instance, the repo is checked out ... | sibyl-oracles/onit | benchmarks/swe_bench_runner.py | .py | fac3c010ad1f857c | 7.48 | 8 |
"""Agentic / tool-use benchmarks (Phase 4 — scaffold).
These are the benchmarks where OnIt's tool loop matters most:
* GAIA — canonical general-assistant, multi-step tool tasks. Run in provider
mode (OnIt drives its own tools); scored by exact match. Uses the gated
``gaia-benchmark/GAIA`` dataset (needs ``HF_TOKE... | sibyl-oracles/onit | benchmarks/tasks/agentic.py | .py | a607d9a70a4a38f0 | 7.48 | 8 |
"""Factuality benchmarks (Phase 3 — scaffold).
Measures short-fact accuracy and hallucination rate. Web-search/fetch tools are
left enabled so the score reflects the *agent*, not just the base model.
Wired in Phase 3: SimpleQA, TruthfulQA, FRAMES. Not yet registered in
``run.py``; add to ``benchmarks/run.py`` TASKS w... | sibyl-oracles/onit | benchmarks/tasks/factuality.py | .py | 8593afb476f6a9b4 | 7.48 | 8 |
"""Reasoning / problem-solving benchmarks.
Phase 1 ships GSM8K as the smoke reasoning task. Later phases add GPQA-Diamond,
MMLU-Pro, MATH/AIME, BBH, and DROP (see the suite plan).
"""
from __future__ import annotations
import re
from inspect_ai import Task, task
from inspect_ai.dataset import Sample, hf_dataset
fro... | sibyl-oracles/onit | benchmarks/tasks/reasoning.py | .py | d4a0c33406f89610 | 7.48 | 8 |
"""Unit tests for the OnIt Inspect provider — no model, network, or Docker.
A stub agent stands in for the real ``OnIt`` so we can verify the provider's
message flattening, the generate() path, and the full Inspect eval+scorer wiring
deterministically.
"""
from __future__ import annotations
import pytest
from inspec... | sibyl-oracles/onit | benchmarks/test_provider.py | .py | 4d4a1ddcdc176939 | 7.98 | 8 |
"""Offline tests for the SWE-bench runner — no Docker, no network, no git."""
from __future__ import annotations
import json
from pathlib import Path
import pytest
from benchmarks import config as bench_config
from benchmarks.swe_bench_runner import (
DATASETS,
_load_existing,
_strip_test_hunks,
_w... | sibyl-oracles/onit | benchmarks/test_swe_bench.py | .py | ca7207ef62648896 | 7.98 | 8 |
"""
# Copyright 2025 Rowel Atienza. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable l... | sibyl-oracles/onit | src/learn/config.py | .py | 6c0e074a89ae5f04 | 7.48 | 8 |
"""
# Copyright 2025 Rowel Atienza. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable l... | sibyl-oracles/onit | src/learn/trajectory.py | .py | 44ddcffa1bdbd10b | 7.48 | 8 |
import os
import shutil
import logging
logger = logging.getLogger(__name__)
# File extensions that indicate code/project files (not session artifacts)
_CODE_EXTENSIONS = {
'.py', '.js', '.ts', '.jsx', '.tsx', '.java', '.c', '.cpp', '.h', '.hpp',
'.cs', '.go', '.rs', '.rb', '.php', '.swift', '.kt', '.scala', '... | sibyl-oracles/onit | src/lib/files.py | .py | 63a0a7d54499dc0f | 7.48 | 8 |
'''
# Copyright 2025 Rowel Atienza. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable l... | sibyl-oracles/onit | src/lib/schema.py | .py | 23258b93bd70224b | 7.48 | 8 |
import re
# Only spans that are actually tag-shaped. A bare "<" is ordinary prose in a
# model answer — "mean 0.709 < 0.712", a generic in a sentence, a stray
# comparison — and a pattern of <[^>]+> pairs that "<" with the next ">" it can
# find, which may be a closing wrapper tag thousands of characters later. The
... | sibyl-oracles/onit | src/lib/text.py | .py | 2f21f58c894bac72 | 7.48 | 8 |
"""Merge config defaults into a user config without overwriting user values."""
from __future__ import annotations
import argparse
import copy
import os
import shutil
import sys
import tempfile
from collections.abc import MutableMapping
from pathlib import Path
from typing import Any
from ruamel.yaml import YAML
fro... | luuquangvu/ha-addons | gemini-fastapi/scripts/merge_config.py | .py | ae40f674364fd916 | 7.66 | 20 |
"""Translate Home Assistant add-on options into the server's environment variables."""
from __future__ import annotations
import json
import os
import sys
from pathlib import Path
from typing import Any
OPTIONS_PATH = Path("/data/options.json")
PROGRAM_MODULE = "wyoming_vietnamese"
# Add-on option name -> environme... | luuquangvu/ha-addons | wyoming-vietnamese/scripts/apply_options.py | .py | 8914267c8f3809cd | 7.66 | 20 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Lightweight global pytest configuration for vLLM-HCU tests.
This module must stay safe to import without initializing an HCU context,
installing runtime patches, loading models, or starting chi... | HYGON-AI/vllm-plugin-das | tests/conftest.py | .py | 3db99a42a12807bb | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""CPU-safe contracts for single-node distributed test declarations."""
from __future__ import annotations
import pytest
from tests.fixtures.distributed import DistributedTopology
@pytest.mark.... | HYGON-AI/vllm-plugin-das | tests/distributed/single_node/test_topology_contracts.py | .py | 8bde9183624559b6 | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Structured metadata shared by hardware and model-test reports."""
from __future__ import annotations
import os
import platform
import sys
from dataclasses import asdict, dataclass, field
from ... | HYGON-AI/vllm-plugin-das | tests/fixtures/artifacts.py | .py | 8bc5fe74bfca1478 | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""CPU-safe HCU resource declarations.
Live torch/HCU probing belongs inside requested fixtures, not module import.
"""
from __future__ import annotations
from dataclasses import dataclass
@da... | HYGON-AI/vllm-plugin-das | tests/fixtures/device.py | .py | e2c268855882493c | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Single-node and reserved multi-node topology declarations."""
from __future__ import annotations
from dataclasses import dataclass
@dataclass(frozen=True)
class DistributedTopology:
tens... | HYGON-AI/vllm-plugin-das | tests/fixtures/distributed.py | .py | c622cc8f57bf1d7e | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""KV-transfer topology and service contracts."""
from __future__ import annotations
from dataclasses import dataclass
from enum import Enum
class TransferScope(str, Enum):
LOCAL_PROCESS = ... | HYGON-AI/vllm-plugin-das | tests/fixtures/kv_transfer.py | .py | ee342126a0613104 | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Stable public contracts for future vLLM-HCU model runners."""
from __future__ import annotations
from dataclasses import dataclass
from typing import Protocol, Sequence
@dataclass(frozen=Tru... | HYGON-AI/vllm-plugin-das | tests/fixtures/model_runner.py | .py | c7de5eac9ec14822 | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Reference-runner contracts for model output comparisons."""
from __future__ import annotations
from typing import Protocol, Sequence
from tests.fixtures.model_runner import GenerationResult
... | HYGON-AI/vllm-plugin-das | tests/fixtures/reference_runner.py | .py | 3553faf2dbd38147 | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Model and dataset resource declarations for integration tests."""
from __future__ import annotations
from dataclasses import dataclass
from pathlib import Path
@dataclass(frozen=True)
class ... | HYGON-AI/vllm-plugin-das | tests/fixtures/resources.py | .py | 730e7f974de56b66 | 8.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Eager versus graph-enabled generation parity on a real small model."""
from __future__ import annotations
import pytest
from tests.fixtures.resources import TestResources as HcuTestResources
f... | HYGON-AI/vllm-plugin-das | tests/integration/graph/test_qwen35_9b_graph_parity.py | .py | baf5f1c90eda026b | 7.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Local KV-transfer connector startup and generation smoke test."""
from __future__ import annotations
from pathlib import Path
import pytest
from tests.fixtures.resources import TestResources ... | HYGON-AI/vllm-plugin-das | tests/integration/kv_transfer/test_example_connector_smoke.py | .py | 3b57117b42ee97c8 | 7.02 | 10 |
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright (c) 2026 Hygon Information Technology Co., Ltd.
"""Single-HCU real checkpoint smoke coverage for the integration layer."""
from __future__ import annotations
import math
from typing import Any
import pytest
from tests.fixtures.resources impor... | HYGON-AI/vllm-plugin-das | tests/integration/models/test_qwen35_9b_smoke.py | .py | 834bb72244947f91 | 8.02 | 10 |
import time
import random
import datetime
import os
import sys
import re
import json
import requests
import pandas as pd
from tkinter import messagebox
class WeiboCommentSpider:
"""微博评论采集模块
负责：
1. 解析微博链接提取帖子ID
2. 采集一级和二级评论
3. CSV输出
"""
def __init__(self, weibo_link_list, max_page, txt_ms... | mashukui/weibo_one_spider | src/weibo_comment.py | .py | 021215f9fffd4787 | 7.45 | 7 |
import time
import random
import datetime
import os
import sys
import re
import json
import requests
import pandas as pd
from bs4 import BeautifulSoup as BS
from tkinter import messagebox
class WeiboSearchSpider:
"""微博搜索采集模块
负责：
1. 关键词搜索微博帖子
2. HTML解析提取帖子信息
3. 图片下载
4. CSV输出
"""
def _... | mashukui/weibo_one_spider | src/weibo_search.py | .py | 9101f8955e345ae9 | 7.45 | 7 |
import time
import random
import datetime
import os
import sys
import re
import json
import requests
import pandas as pd
from tkinter import messagebox
class WeiboUserPosted:
"""微博用户主页帖子采集模块
负责：
1. 遍历用户链接列表
2. 分页获取用户发布的微博列表
3. 支持关键词/时间筛选
4. 长微博文本展开
5. CSV输出
"""
def __init__(self,... | mashukui/weibo_one_spider | src/weibo_user_post.py | .py | a136553d56b65e30 | 7.45 | 7 |
import time
import random
import datetime
import os
import sys
import csv
import re
import json
import requests
from tkinter import messagebox
GENDER_MAP = {'m': '男', 'f': '女', 'n': '未知'}
VERIFIED_TYPE_MAP = {
0: '未认证', 1: '个人认证', 2: '企业认证', 3: '媒体认证',
4: '政府认证', 5: '校园认证', 6: '机构认证', 7: '其他认证',
}
def _par... | mashukui/weibo_one_spider | src/weibo_userinfo.py | .py | 376b299460c1f930 | 7.45 | 7 |
"""romp_colormap — the recency colormaps shared by every romp view.
Single source of truth: romp-feed, the kernel, and the render bundle (and any future
view / tmux segment) import age_rgb from here, so the look is ONE edit, not N.
age_rgb(age_seconds, name) -> (r,g,b): recency on a LOG scale — most recent maps to th... | romp-on/romp | kernel/colormap.py | .py | 0708626d126347f7 | 7.6 | 15 |
"""romp_palette — the selectable session-identity palettes (categorical).
The identity palette is the set session tabs, feed cards, timeline lanes, and
postal bubbles draw their per-session color from. It used to be ONE hardcoded
9-color list copied across three assigners (bin/romp, the kernel, the SDK
backend); the S... | romp-on/romp | kernel/palette.py | .py | 15dbc898628a1a93 | 7.6 | 15 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.