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/specsnl/agent-specs/code-commentsnpx skills add specsnl/agent-specs --skill code-commentsgit clone --depth 1 https://github.com/specsnl/agent-specsWhat 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 | $0.00145 | $0.01401 |
| Opus 5 | $0.00072 | $0.00700 |
| Sonnet 5 | $0.00029 | $0.00280 |
| Haiku 4.5 | $0.00015 | $0.00140 |
Grade A, and why
code-comments 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 yesterday.
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 — 167 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Code comments
The default: write no comment
Code plus good naming is the primary form of explanation. A comment is an exception that must earn its place by carrying information a competent reader cannot recover from the code itself.
If you are about to write a comment, the burden is on the comment. Do not write one unless it falls under Warranted below.
Match the surrounding file first
Local convention beats this skill's defaults, always. If the file — or its neighbours — documents every public method, keep documenting them in the same shape. This skill is never a reason to strip comments from a heavily-documented codebase, nor to introduce a documentation style the repo does not use.
Scope
This governs the comments you write. It is not a mandate to delete existing comments in files you happen to touch, and it is not a docs-generation skill. The one exception: if you change code such that an adjacent comment becomes wrong, update or delete that comment — a false comment is worse than none.
Warranted
Write a comment only when it is one of these:
- Why, not what. A non-obvious decision, tradeoff, or constraint.
// Cast: the upstream API returns numeric strings for these fields. $total = (int) $payload['total']; - A workaround and its cause, ideally with a link — an upstream bug, a vendor quirk, a spec
deviation.
// Safari fires resize before layout settles; see https://bugs.webkit.org/show_bug.cgi?id=170595 requestAnimationFrame(measure); - Non-obvious consequences. Ordering that matters, a lock that must be held, a side effect a
caller would not expect.
// Must run before Close(): the flush path reads the connection. w.Flush() - Genuinely subtle algorithms or math, or a regex that is not self-evident.
- Public API contracts, where the language's ecosystem expects them and the file already has them.
TODO/FIXMEonly when the user asked for it or agreed to defer the work. Never a self-issued note about something you skipped — either do it or tell the user in your response.
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.
- yesterday First seen · 167 lines · 145 tokens per session scan A 8a1f693d2093
code-comments is a skill published in the GitHub repository specsnl/agent-specs (2 stars, last pushed 16d ago), licensed MIT. It adds 145 tokens to every session and 1,401 once invoked, about $0.0007 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
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…