菜单
Generate asset-actions.yaml from ASSETS.yaml by classifying assets into priority tiers. Use when the user asks to regenerate, update, or refresh the asset actions.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/UKGovernmentBEIS/inspect_evals --skill generate-asset-actions命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | generate-asset-actions |
| description | Generate asset-actions.yaml from ASSETS.yaml by classifying assets into priority tiers. Use when the user asks to regenerate, update, or refresh the asset actions. |
Regenerate internal/audits/asset-actions.yaml and internal/audits/audit-summary.md from ASSETS.yaml.
If ASSETS.yaml may be stale, run uv run python tools/generate_asset_manifest.py first.
Run uv run python tools/summarise_asset_manifest.py to get aggregate counts (by type, by state, totals). Use these numbers when populating audit-summary.md.
Read ASSETS.yaml. For each asset, determine target stage first, then priority. Process both state: floating assets AND state: pinned assets that match known-unstable sources (since their target is controlled, they are not yet at their target stage).
The target stage depends on host reliability, not asset type:
controlled (Stage 2) — any asset where upstream has broken before, maintainer is unresponsive/deprecated, OR host is unreliable (personal repos, Google Drive, .edu domains, university servers, any host without version control). This applies to git_clone, direct_url, and huggingface alike.pinned (Stage 1) — assets on reliable, version-controlled hosts (GitHub, HuggingFace, well-known CDNs) with no history of breakage.Per ADR-0007: "Anything hosted on a less reliable domain (personal websites, Google Drive, university servers, or any host without version control) should skip straight to Stage 2."
pinned.controlled.drive.google.com, .edu domains, personal repos/websites) not already in the known-unstable registry. Target is controlled.For assets with state: pinned and a {SHA} placeholder but no checksum, classify as Low (target: pinned with checksum).
Omit assets already at their target stage.
Every entry needs: eval, source, type, state, target, action, reason.
Update this list when new instability is discovered.
| Source | Eval | Incident |
|---|---|---|
xlang-ai/OSWorld | osworld | Files removed (PR #958) |
openai/evals | makemesay | Deprecated upstream |
corebench.cs.princeton.edu | core_bench | University server, no versioning |
epatey/fonts | osworld | Personal repo |
ShishirPatil/gorilla | bfcl | Data format issues (PR #954) |
yunx-z/MLRC-Bench | mlrc_bench | Broken task |
LRudL/sad | sad | Upstream bugs (issues #7, #8) |
meg-tong/sycophancy-eval | sycophancy | Invalid JSON/NaN, workaround in code |
josancamon/paperbench | paperbench | Paper ID mismatch (HF discussion #2) |
sentientfutures/moru-benchmark | moru | Exact duplicate rows |
asset-actions.yaml parses as valid YAMLfloating_assets + needing_checksums + no_action_needed == total_external_assetsaudit-summary.md match output of summarise_asset_manifest.pyCrawl repository PRs, issues, and review comments to distill institutional knowledge into a shared knowledge base. Run periodically by "context agents" to maintain agent_artefacts/repo_context/REPO_CONTEXT.md. Trigger only on specific request.
Use Inspect Scout to analyze agent trajectories from evaluation log files. Runs default and custom scanners to detect external failures, formatting issues, reward hacking, and ethical refusals. Use when user asks to check/analyze agent trajectories. Trigger when the user asks you to run the "Check Agent Trajectories" workflow.
CI and GitHub Actions maintenance workflows — fix a failing test from a CI URL, fix a failing smoke test, add @pytest.mark.slow markers to slow tests, or review a PR against agent-checkable standards. Use when user asks to fix a failing test, fix a smoke test, mark slow tests, or review a PR. Trigger when the user asks you to run the "Write a PR For A Failing Test", "Fix A Failing Smoke Test", "Mark Slow Tests", or "Review PR According to Agent-Checkable Standards" workflow.