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/kkollsga/codingest/releasenpx skills add kkollsga/codingest --skill releasegit clone --depth 1 https://github.com/kkollsga/codingestWhat 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 | $0.00117 | $0.07784 |
| Opus 5 | $0.00059 | $0.03892 |
| Sonnet 5 | $0.00023 | $0.01557 |
| Haiku 4.5 | $0.00012 | $0.00778 |
Grade A, and why
release scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
(`curl -s -H "User-Agent: codingest-release" https://crates.io/api/v1/crates/<c> | jq -r .crate.max_version` How it starts
The opening of the file, as written. The whole thing — 450 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Release
This run ends in a published version or a named blocker
Read this before step 1; it governs every step below (doctrine [[R12]]).
The completion condition. The run is done when step 11 has verified
x.y.z on crates.io (all three crates) and on PyPI with the full artifact set,
and the GitHub Release exists and is not a draft — or when you have surfaced a
specific blocker you cannot fix. Nothing else is an ending.
The non-endings. "CI is running", "the commit is staged", "the tag is pushed", "waiting for the wheels", "next I will…" are not endings. Each is a natural pause point that feels like a reasonable place to hand back control, and each is indistinguishable from the inside from genuinely needing input. The failure is not a decision to stop — it is failing to notice that continuing is an option. Three releases across the estate stalled exactly this way on 2026-07-30/31, every check green, no version shipped; the user noticed before the agent did, twice.
Waiting is not a checkpoint. Poll it, background it, or block on it — all three continue the run. Handing control back does not. A poll that takes twenty minutes is twenty minutes of polling, not a reason to end the turn. The crates.io index-propagation wait and the multi-platform wheel builds are the two places this run is most likely to stall.
A red CI is a task, not a verdict. Diagnose, fix, push, re-poll, repeat — within step 10's stated bound (~3 iterations), then surface what remains as a named blocker.
Preconditions
- No double-stage. Check no release is already staged:
git log origin/main..HEAD --oneline | grep -E "release\(". If arelease(x.y.z)sits unpushed, keep that version — fold work into the same[x.y.z]block (one version bump per release). - Surgical staging. If there's unrelated uncommitted work in the tree, don't
block on it and don't sweep it in: stage every release file explicitly by
path (
git add <file> …, nevergit add -A/.) and verify withgit status --porcelainthat only release files are staged. Leave unrelated changes untouched for their author.- That verification is not a formality:
git addis all-or-nothing across its pathspecs. One bad path — a typo, a file the plan renamed, aBENCHMARKS.mdyou did not actually refresh — aborts the entire invocation, so none of the other files are staged either. The failure is quiet in the reassuring direction: the nextgit commitsucceeds, on a release commit missing the bump. Read back the staged set (git diff --cached --name-only) and confirm it matches the intended list before committing. (KGLite hit this on 2026-08-09.)
- That verification is not a formality:
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 · 450 lines · 117 tokens per session scan A 33bb5ad8e9bb
release is a skill published in the GitHub repository kkollsga/codingest (0 stars, last pushed yesterday), licensed MIT. It adds 117 tokens to every session and 7,784 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). 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…