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 shimo4228/claude-harness --skill learn-evalgit clone --depth 1 https://github.com/shimo4228/claude-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/shimo4228/claude-harness/learn-eval)<a href="https://agentmods.dev/skills/shimo4228/claude-harness/learn-eval"><img src="https://agentmods.dev/badge/skills/shimo4228/claude-harness/learn-eval/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/shimo4228/claude-harness/learn-eval"><img src="https://agentmods.dev/badge/skills/shimo4228/claude-harness/learn-eval.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 | $0.00042 | $0.03552 |
| Opus 5.5 | $0.00017 | $0.01421 |
| Sonnet 5.5 | $0.00008 | $0.00710 |
| Haiku 4.5 | $0.00004 | $0.00355 |
Grade A, and why
learn-eval 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 10d 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 — 261 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/learn-eval - Extract, Evaluate, then Save
Extract a reusable pattern from the session, gate it, and route every Save to a destination something actually reaches.
What to Extract
Look for:
- Error Resolution Patterns — root cause + fix + reusability
- Debugging Techniques — non-obvious steps, tool combinations
- Workarounds — library quirks, API limitations, version-specific fixes
- Project-Specific Patterns — conventions, architecture decisions, integration patterns
Process
-
Review the session for extractable patterns
-
Identify the most valuable/reusable insight
-
Determine the destination — there is no parking lot.
Every Save must land somewhere that something actually routes to (ADR-0047), so pick one of exactly two:
- Absorb into an existing asset — the pattern belongs inside a skill, rule, or
hooks/README.mdsection that already owns the topic. Name the file and the section. This is the default: an addition to a reachable asset beats a new file. A skill's value lies in its comprehensiveness, so appending improves it — "I don't want to pollute the existing skill" routes the pattern to a file nothing reaches. - Promote to a skill — the pattern has its own independent trigger (a user request
that no installed skill answers). Run skill: skill-creator (required by
rules/common/skills.mdbefore writing any skill).
If neither fits, the verdict is Drop, not "park it somewhere for now". A note that nothing points at is reachable only by grep, and grep requires already knowing the content exists (measurement: ADR-0047).
Global vs project placement (once a destination type is chosen): the source of truth is
docs/adr/0025-global-vs-project-asset-placement.md. - Absorb into an existing asset — the pattern belongs inside a skill, rule, or
-
Draft the candidate as a scratch note (the final skill shape belongs to
skill-creator;overlap_candidates.pyreads name / description / Problem / Solution / When to Use):
What ships with it
5 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.
- 10d ago Changed · +3 lines 6920b7f764be
- 11d ago Changed · -94 tokens per session 8411fe883c0a
- 13d ago Changed · -1 lines 40c88fab2d86
- 24d ago Changed · -1 tokens per session 53a138b21501
- 25d ago Changed · -2 lines 23c0a48d47ca
- 1mo ago Changed · +7 lines c1de986d7ac5
- 1mo ago Changed · -10 lines c895e843d422
- 1mo ago First seen · 264 lines · 137 tokens per session scan A 931adb1360ce
learn-eval is a skill published in the GitHub repository shimo4228/claude-harness (3 stars, last pushed 2d ago), licensed MIT. It adds 42 tokens to every session and 3,552 once invoked, about $0.0002 per session on Opus 5.5. 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
agent-adoption-triage
Five-question triage for routing AI work to the right architecture — script, algorithmic search, LLM workflow, or autonomous agentic loop — before any agent is built. Use when someone says "let's build an agent for X", when deciding whether a use case actually needs an autonomous agent, when reviewing an agent…
git-commit
A guided Git commit workflow that examines changes and creates a commit message using the Conventional Commits format, a shared style for labeling changes such as features, fixes, tests, or documentation.
zod
Zod v4 best practices, patterns, and API guidance for schema validation, parsing, error handling, and type inference in TypeScript applications. Covers safeParse, object composition, refinements, transforms, codecs, branded types, v3→v4 migration, and testing schemas with Jest or Vitest. Baseline: zod ^4.6.0. Triggers…
pi-triage
Decide whether an upstream pi change needs porting to the Go port. Use when assessing upstream commits/PRs ("should we port X?"), or as the triage stage of /pi-sync. Outputs a WHY/WHAT/SCOPE verdict per change.
trailhead-work
Skill "trailhead-work" from ToRvaLDz/trailhead, covering load first, in order, routing: verb to engine, work the map: work, work one ticket whole, without splitting: quick and pause, resume, split.
pi-parity-review
Adversarially verify that a ported change is faithful to the original pi implementation (TS source + published npm build). Use after porting upstream pi changes, or standalone on any area of this repo ("is X faithful to pi?").