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 enmanuelmag/agent-harness-kit --skill ahk-reviewgit clone --depth 1 https://github.com/enmanuelmag/agent-harness-kitWrote 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/enmanuelmag/agent-harness-kit/ahk-review)<a href="https://agentmods.dev/skills/enmanuelmag/agent-harness-kit/ahk-review"><img src="https://agentmods.dev/badge/skills/enmanuelmag/agent-harness-kit/ahk-review/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/enmanuelmag/agent-harness-kit/ahk-review"><img src="https://agentmods.dev/badge/skills/enmanuelmag/agent-harness-kit/ahk-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 28 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 46 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
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 | $0.00035 | $0.00717 |
| Opus 5 | $0.00017 | $0.00358 |
| Sonnet 5 | $0.00007 | $0.00143 |
| Haiku 4.5 | $0.00003 | $0.00072 |
Grade A, and why
ahk-review 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.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Provider Delegation Guidance
- Sequential: Delegate only bounded, independent work; retain decisions and the final synthesis in the parent thread.
- Parallel: Delegate independent work in parallel when useful, then wait for every delegated result before continuing.
- Context Transfer: Give every delegated task a self-contained objective, scope, relevant context, restrictions, and output contract.
- Wait For Completion: Wait for the delegated result, then consolidate its findings in the parent thread.
- Inspect Progress: In the interactive CLI, use /agent to inspect delegated threads when needed.
You are in lightweight review mode. The ticket/objective: $ARGUMENTS
If
$ARGUMENTSis empty, ask the user to provide the ticket guidelines or objective before doing anything else.
Rules
- NO MCP calls — no tasks., no actions., no tasks.acceptance.update, no other MCP write of any kind
- NO health.sh unless explicitly relevant to the objective
- NO builder, NO reviewer
- Create files ONLY if user explicitly asked to save the output
- This is a preview/rubric generator, not the official reviewer — it never mutates harness state
Process
-
Proactively enumerate the skills currently available/installed in your context and decide which are relevant to the objective. The provider has already injected available skills into your context — check what skills you are aware of, identify which are relevant to this objective, and reference them explicitly in your review. If no installed skills seem relevant to the objective, recommend running
npx autoskillsto fetch appropriate skill packs. -
Invoke Explorer as a subagent with this exact instruction:
"Read-only review investigation — no MCP harness, no task creation. Ticket/objective:
$ARGUMENTS. Map the files relevant to this objective — what changed, what's implicated, what patterns are involved. Additionally, for each relevant function or module you touch, read its full body and evaluate whether its name matches what it actually does — flag any name/behavior mismatch with an exact file:line reference, even if the mismatch falls outside the direct scope of the stated objective. Return both the file map and the name/behavior audit as plain structured text."
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 Changed · +9 lines 5405c6f28e14
- 12d ago First seen · 46 lines · 35 tokens per session scan A 8819d521578c
ahk-review is a skill published in the GitHub repository enmanuelmag/agent-harness-kit (181 stars, last pushed today), licensed Apache-2.0. It adds 35 tokens to every session and 717 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
OpenCode
Frames OpenCode as an operator-controlled coding lane for bounded implementation or review work inside a real repository.
review-prs
Review a GitHub pull request in the googleapis/mcp-toolbox repo against the team's reviewer checklist: PR title/description conventions, linked issue, logic errors and unhandled edge cases, breaking changes, test coverage, docs updates, security (input handling), and new dependencies. Use whenever a maintainer asks…
bootstrap-repo-analysis
First-time analysis of a repository with no prior reviewer outcomes. Crawl historical merged-PR review feedback with the gh CLI (plus any preloaded samples), extract the team's review norms, and synthesize the initial per-repo review-style prompt. Use this for a cold-start repo; use continual-learning instead once the…
continual-learning
Nightly refinement of an existing per-repo review-style prompt using this reviewer's own finding outcomes. Read confirmed (resolved-by-commit / thumbs-up) and dismissed (thumbs-down) findings, promote the bug patterns the team actually fixes, demote the false-positive patterns, reconcile against the current prompt…
github
GitHub operations via gh CLI: issues, PRs, CI runs, code review, API queries. Use when: (1) checking PR status or CI, (2) creating/commenting on issues, (3) listing/filtering PRs or issues, (4) viewing run logs. NOT for: complex web UI interactions requiring manual browser flows (use browser tooling when available)…
meta-codereview-current-diff
Read the current uncommitted diff, run three independent reviewers (safety + tests-coverage + style) in parallel, then arbitrate a single BLOCK / BLOCKWITHOVERRIDE / PASSWITHNOTES verdict. Use before commit when you want a multi-perspective second-opinion instead of a single-reviewer agent loop.