Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/redker56/auto-harness/evaluator-write-qanpx skills add redker56/auto-harness --skill evaluator-write-qagit clone --depth 1 https://github.com/redker56/auto-harnessWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/redker56/auto-harness/evaluator-write-qa)<a href="https://agentmods.dev/skills/redker56/auto-harness/evaluator-write-qa"><img src="https://agentmods.dev/badge/skills/redker56/auto-harness/evaluator-write-qa.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00032 | $0.00639 |
| Opus 5 | $0.00016 | $0.00319 |
| Sonnet 5 | $0.00006 | $0.00128 |
| Haiku 4.5 | $0.00003 | $0.00064 |
Grade A, and why
evaluator-write-qa scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Evaluator Write QA
This skill governs Evaluator qa mode and the sprint QA report.
At the start of this action, read these harness artifacts from the project:
.harness/status.md.harness/intake.md.harness/spec.md.harness/design-direction.md.harness/runtime.md.harness/contracts/sprint-XX-contract.md.harness/contracts/sprint-XX-review.mdwhen a review exists for the sprint..harness/qa/sprint-XX-self-check.mdwhen Generator produced a self-check..harness/qa/sprint-XX-qa-report.mdwhen rewriting the QA report.
Then inspect the current project implementation relevant to this action:
- Read the source files, routes, components, services, tests, and config that implement the sprint you are validating.
- Use running-app evidence first, and use the codebase to explain, verify, or localize what you observe.
Then read these skill references:
references/protocols/file-ownership.mdreferences/templates/qa-report.mdreferences/rubrics/default-grading.mdreferences/rubrics/product-depth.mdreferences/rubrics/bug-severity.mdreferences/rubrics/visual-design.mdreferences/rubrics/code-quality.md
Follow these rules:
- Before issuing PASS or FAIL, complete the main end-to-end user flow promised by the current sprint contract against the running app.
- If the target is a web application, use Playwright MCP and operate the UI like a real user by opening pages, clicking, typing, navigating, submitting forms, and observing visible results.
- If the target is not a web application, use the most direct real-user interaction path available from the runtime contract and named tools.
- If you cannot complete the main contracted user flow from the running app, do not pass on file inspection alone; record the verification gap and fail conservatively.
- Use the QA report template exactly.
- Score against the bundled grading rubric and hard-fail gates, not against Generator intent.
- Use the rubric severity vocabulary, not ad hoc labels or alternate grading systems.
- Include a
Deduction Ledgerthat traces every actual deduction or hard-fail trigger used to derive the final scorecard. - Make the
Scorecardderivable from the rubric baselines and the rows inDeduction Ledger. - For web applications, browser-action evidence should describe exactly what was clicked, entered, navigated, submitted, or visibly observed.
- Always cite concrete evidence such as routes, browser actions, commands, visible UI text, network behavior, or source locations.
- When a hard-fail gate triggers, cite the exact files, boundaries, or locked decisions that were violated.
- Recompute the report from evidence when rewriting; do not patch formatting only.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 53 lines · 32 tokens per session scan A 8eeea9bdab09
evaluator-write-qa is a skill published in the GitHub repository redker56/auto-harness (14 stars, last pushed 4mo ago), licensed MIT. It adds 32 tokens to every session and 639 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
harness-creator
Build, audit, and improve harnesses that make AI coding agents reliable: AGENTS.md/CLAUDE.md instruction files, feature/state tracking, verification gates, scope boundaries, session handoff, memory persistence, context budgets, tool-permission safety, and multi-agent coordination. Use this whenever a coding agent is…
harness
하네스를 구성합니다. 전문 에이전트를 정의하며, 해당 에이전트가 사용할 스킬을 생성하는 메타 스킬. (1) '하네스 구성해줘', '하네스 구축해줘' 요청 시, (2) '하네스 설계', '하네스 엔지니어링' 요청 시, (3) 새로운 도메인/프로젝트에 대한 하네스 기반 자동화 체계를 구축할 때, (4) 하네스 구성을 재구성하거나 확장할 때, (5) '하네스 점검', '하네스 감사', '하네스 현황', '에이전트/스킬 동기화' 등 기존 하네스 운영/유지보수 요청 시 사용.
push-ci
Push to remote and monitor CI. Validates branch safety, executes git push WITH explicit user approval, then monitors CI run status via gh CLI. Use when: user says 'push', 'push and watch CI', 'ship it', 'push-ci'. Not for: committing (use /smart-commit), creating PRs (use /create-pr), merging (use /merge-prep).
create-request
Create, update, or scan per-task request tickets for progress tracking. These are date-prefixed non-lifecycle docs under requests/, NOT feature-level requirements (use /req-analyze for those). Use when: tracking task progress, updating completion status, scanning incomplete requests, checking request status dashboard.…
smart-rebase
Smart partial rebase for squash-merge repositories. Auto-detect which commits to keep/drop when base branch was squash-merged into target. Use when: user says 'rebase', 'partial rebase', 'base already merged', 'smart rebase', or /smart-rebase. Not for: simple git rebase (the developer runs it — Claude never executes…
remind
Lightweight model correction with context-aware rule loading. Use when: model forgot a rule, skipped a required step, edited code/docs without running review, needs to re-read CLAUDE.md or rules. Triggers on: 'you forgot', 'remind', 'check rules', 'what did you miss', '你忘了', 'did you skip review', 'why didn't you run…