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 skills add yfge/agent-harness-skills --skill repo-harness-assessmentgit clone --depth 1 https://github.com/yfge/agent-harness-skillsWrote 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/yfge/agent-harness-skills/repo-harness-assessment)<a href="https://agentmods.dev/skills/yfge/agent-harness-skills/repo-harness-assessment"><img src="https://agentmods.dev/badge/skills/yfge/agent-harness-skills/repo-harness-assessment/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/yfge/agent-harness-skills/repo-harness-assessment"><img src="https://agentmods.dev/badge/skills/yfge/agent-harness-skills/repo-harness-assessment.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00056 | $0.01093 |
| Opus 5 | $0.00028 | $0.00547 |
| Sonnet 5 | $0.00011 | $0.00219 |
| Haiku 4.5 | $0.00006 | $0.00109 |
Grade A, and why
repo-harness-assessment 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 12d 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.
How it starts
The opening of the file, as written. The whole thing — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repo Harness Assessment
Overview
Assess how well a repository lets an agent find its rules, make safe changes, verify them, and produce reviewable evidence.
This is the default router for existing repositories. It also owns agent-entrypoint design because entrypoints are the navigation layer of the assessment, not a separate harness system. For shared vocabulary and neutral artifact names, see ../../references/harness-patterns.md; when expected harness files are missing, use references/build-when-missing.md; for canonical entrypoints, mirrors, and drift prevention, use references/entrypoint-policy.md.
When To Use
- The user asks what harness pieces a repository is missing.
- The user wants to create, shrink, reconcile, or diagnose
AGENTS.md,CLAUDE.md,GEMINI.md, Cursor rules, or GitHub instructions. - You need to compare entrypoints, validation commands, runtime evidence, delivery records, or quality gates across repositories.
- You need to decide whether the next smallest improvement is an entrypoint, validation script, artifact bundle, ledger, contract check, or quality gate.
Inputs Needed
- Repository root path.
- User scope: whole repository, one surface, docs-only work, runtime behavior, CI, or delivery flow.
- Any expected harness shape or maturity target the user names.
- Existing agent instruction files and whether mirrors, generation, or subtree overrides are required.
Execution Order
- First: Read repository entrypoints and source-of-truth files, including agent instructions, README, architecture or reliability docs, indexes, CI, and scripts.
- Then: Map existing surfaces to harness roles and check entrypoint precedence, mirrors, validation, evidence, work state, delivery, contracts, and quality.
- Finally: Report maturity, entrypoint actions, the smallest useful improvement slice, and what not to build yet.
Step-by-Step Process
- Use
rg --filesorfindto list agent instruction files, docs, scripts, CI, work-state surfaces, ledgers, reports, and runtime artifacts. - Identify one canonical entrypoint and classify every other agent instruction file as a subtree override, symlink, generated mirror, or short pointer.
- Keep the root entrypoint to scope, source-of-truth navigation, hard boundaries, and minimum commands; move detailed procedures to linked docs.
- Map current artifacts to entrypoint, work-state, ledger, contracts, validation, runtime-evidence, and quality roles before proposing new files.
- Check for a stable validation matrix and whether failures connect to run IDs, request IDs, logs, screenshots, JSON/JUnit output, reviews, or commits.
- If a required role is absent, define the minimum bootstrap artifact from
references/build-when-missing.md; do not scaffold optional roles by default. - Add a mirror or pointer drift check when multiple agent instruction surfaces must stay aligned.
- Compress gaps into no more than three next steps, ordered by value and risk.
What ships with it
2 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.
- 12d ago First seen · 113 lines · 56 tokens per session scan A fbd6635ae94a
repo-harness-assessment is a skill published in the GitHub repository yfge/agent-harness-skills (10 stars, last pushed 1mo ago), licensed MIT. It adds 56 tokens to every session and 1,093 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
readonly-review
Run a no-network read-only review practice with two explorer steps and one synthesis step.
coding-guardrails
Unified LLM coding rules, simplicity guidelines, surgical changes, and thinking protocols for all AI agents. Triggers on: code rules, guardrails, coding guidelines, coding conventions, coding style, surgical changes, goal-driven.
breaking-down-the-work
Splits scope into a product-first work breakdown that follows the 100% rule, keeps pieces from overlapping, uses outline numbers, and gives every piece a dictionary entry. Use when an epic, feature, or new subsystem needs a clean split into deliverables, or one source of truth before folders or work begin. Do not use…
staying-on-mission
Tests the current work against a written, lasting mission anchor and forces one choice, re-anchor, escalate, or stop. Use when an agent keeps finishing tasks but the work drifts from the goal, scope creeps, the same action repeats in a loop, or standards slip one small step at a time. Do not use for a tiny edit with…
proving-claims
Builds claim-to-evidence trace rows with statuses, gaps, tests, evals, reviews, and narrower non-claims. Use when a packet asserts something a reviewer must trust. Do not use to make the final ship decision, judge public legal wording, or invent evidence that does not exist.
responding-to-incidents
Runs a live incident the stabilize-first way — name a commander, separate facts from hypotheses, prefer reversible actions, communicate on a cadence, and drive corrective actions to closure. Use when production is broken, data is at risk, or an agent action caused harm. Do not use for routine non-incident work, or as…