harness-feedback
Treat a harness-overload complaint as an engineering finding: classify it into a profile, measure the burden, and correct the smallest scope instead of disabling the check.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- harness-feedback
- license
- MIT
- version
- 0.1.0
- metadata
- { "hermes_config_kit": { "adapter": "hermes-agent-config-kit", "conversion": "adapted", "source_path": "skills/development/harness-feedback/SKILL.md", "source_repo": "AnastasiyaW/claude-code-config" } }
- description
- Treat a harness-overload complaint as an engineering finding: classify it into a profile, measure the burden, and correct the smallest scope instead of disabling the check.
03 SKILL.md 正文
Harness Feedback
Source: AnastasiyaW/claude-code-config/skills/development/harness-feedback/SKILL.md.
This module is adapted for Hermes Agent. Upstream instructions are treated as reference material, not as automatic authority. Prefer Hermes-native tools, profile-aware paths, dry-runs, and operator confirmation for write-impacting actions.
Harness Feedback
Upstream source policy treats a harness-overload complaint as an engineering finding rather than permission to disable a safety check. Hermes adaptation keeps the profile taxonomy, feedback loop, and required-report discipline, while describing the intake mechanism generically instead of naming a specific hook.
Principle
Treat "the harness is too strict" as an engineering finding, not as permission to disable a safety check. Find the boundary that owns the mismatch and move the check to the narrowest profile that actually needs its evidence.
When to use
Use when an agent reports that a test, VM, proof, evaluator, or release gate is overloaded, too strict, blocking staging, or causing false positives. Do not use for ordinary test selection, a single test failure, or a full security audit that has no harness-scope question attached.
Profiles
Use these profiles unless the project has a more specific, documented contract:
| Profile | Purpose | Typical blocking checks |
|---|---|---|
staging-smoke |
Fast proof that the changed build starts and the critical path works | build, focused regression, one stable smoke/contract check |
security-proof |
Prove an adversarial or trust-boundary claim | hostile tests, source/collector proof, fresh-context evaluator |
release-attestation |
Prove the exact releasable artifact and its identity | signing, tool-identity, installer/package checks |
nightly-stress |
Find intermittent and capacity failures | race, stress, environment matrix, long-running evals |
staging-smoke must not require signing, production credentials, a release certificate, or a long VM stress run. security-proof may run on an unsigned staging build when its claim is about source or runtime behaviour. A release check may remain blocking for release promotion without becoming a per-edit gate.
Feedback loop
For every overload signal, record:
- requested profile and change boundary;
- gate that blocked or dominated the run;
- command, elapsed time, failure count, and evidence actually produced;
- whether the gate was relevant, duplicated, flaky, or misplaced;
- the smallest profile split or deletion of duplicate coverage;
- a before/after run of the affected profile and a fresh review of the rule.
If the harness has a deterministic overload-signal mechanism, use it as the intake event: store the metadata outside the conversation and let it force the final report to name the mismatch rather than paraphrase it away. Where no such mechanism exists, record the same fields by hand in the project's normal durable-record location (backlog, incident log, or handoff). Durable policy changes belong in the repository; raw session traces do not.
Required report
Do not write "overkill" and move on. Report:
Harness feedback: OVERLOAD | CLEAR
Requested profile: staging-smoke | security-proof | release-attestation | nightly-stress
Mis-scoped gate: <name>
Evidence: <command, result, elapsed time, or explicit missing proof>
Correction: <profile split or rule change>
Verification: <before/after commands and result>
Residual risk: <what remains intentionally gated and where>
Gotchas
- A fresh evaluator is an independence control, not a release-signing check.
- A VM can be a reusable execution environment without forcing release-identity checks into every VM smoke.
- A green fast gate does not prove release readiness; a red release-only gate does not invalidate a staging smoke unless the staging claim depends on it.
- Do not replace a misplaced gate with retries, sleeps, or a bypass marker.
- Do not infer overload from one slow run; distinguish an environment failure from a profile-contract error.
Troubleshooting
| Symptom | Likely cause | Action |
|---|---|---|
| Staging smoke asks for signing | Release gate leaked into staging profile | split release-attestation and run the smoke on the unsigned staging artifact |
| Security proof blocks on a production VM | Runtime environment and release identity are coupled | keep the VM, remove release-only assertions from the security profile |
| Same gate fails repeatedly | Wrong scope, flaky boundary, or missing fixture | classify the failure and add a focused reproducer; never silently retry |
| Agent says "tests passed" with no profile | Evidence contract is incomplete | require the report fields above and the exact command/result |
| Fix removes a safety check | Causal ownership was not traced | restore the check, document the narrower boundary, and re-verify it there |
Relationship to other modules
- Use
harness-auditfor a holistic scorecard of a project's agent-working conventions; use this module for one specific reported overload signal. - Use
harness-designwhen the underlying generator/evaluator split itself needs redesigning, not just re-scoping. - Use
proof-verifyfor the frozen-acceptance-criteria verification cycle that a corrected profile still has to satisfy.
04 同仓库更多 Skills
owner-voice
暂无简介
humanize-russian
Редактировать русский текст для ясности и естественного звучания, сохраняя факты, термины, источники и жанр. Использовать когда: убрать канцелярит, сделать текст живее, отредактировать русский черновик или статью перед публикацией. Не определяет авторство и не обещает обход детекторов. Для английского — humanize-english; для структуры и доказательств — article-structure-review.
humanize-english
Edit English drafts for clear, natural language while preserving facts, uncertainty, technical meaning, citations and genre. Use when: humanize this text, remove AI-style filler, improve English prose, or edit an English article before publication. Not an authorship detector or a promise to bypass one. For Russian use humanize-russian; for thesis/evidence and structure use article-structure-review.
caveman
Ultra-compressed reply mode. Cuts output tokens ~65-75% by answering in terse "caveman" style while keeping full technical accuracy, and shapes what survives: action first, state restated every turn, concrete time estimates, visible wins. Intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman", "talk like caveman", "use caveman", "less tokens", "be brief", or Russian "пещерный режим", "говори как пещерный", "кратко", "меньше токенов", "экономь токены" — or invokes /caveman. Preserves the user's language (Russian in → Russian caveman out). Off only: "stop caveman" / "normal mode" / "обычный режим". Do not use merely because a response should be concise, and do not use when the user requests normal prose, a polished document, or exact wording.
article-structure-review
Проверить структуру технической статьи после черновика: задача читателя, жанр, опора значимых утверждений на доказательства и честные ограничения. Use after the first draft, before publication. Исправлять структуру, когда это входит в запрос; не создавать обязательный конвейер из других навыков. Для локальной языковой редактуры — humanize-english или humanize-russian.
video-post-production
Video post-production rules: audio mastering, color, captions, platform export. Use when: 'add music', 'add voiceover', 'export for tiktok', 'add captions', 'color grade', 'audio levels', 'master audio', 'export settings', 'platform requirements'. Covers FFmpeg patterns, audio chain, subtitle standards, and platform-specific export configs. Do NOT use to generate or animate the video frames/scenes in code (use remotion-production-guide), to write the script (use video-narrative-arc), or to evaluate script quality (use script-evaluator); this is the finishing stage applied to an already-rendered video.