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 qinghaoliqaq/AutoCoder --skill implement_specsgit clone --depth 1 https://github.com/qinghaoliqaq/AutoCoderWrote 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/qinghaoliqaq/autocoder/implement_specs)<a href="https://agentmods.dev/skills/qinghaoliqaq/autocoder/implement_specs"><img src="https://agentmods.dev/badge/skills/qinghaoliqaq/autocoder/implement_specs/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/qinghaoliqaq/autocoder/implement_specs"><img src="https://agentmods.dev/badge/skills/qinghaoliqaq/autocoder/implement_specs.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 | $0.00059 | $0.01129 |
| Opus 5.5 | $0.00024 | $0.00452 |
| Sonnet 5.5 | $0.00012 | $0.00226 |
| Haiku 4.5 | $0.00006 | $0.00113 |
Grade A, and why
implement-specs 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 1mo 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 — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Implement Specs
This skill turns a specs/<id>/TECH.md into committed code. Its core
discipline:
The spec and the code update in the same PR. A drifted spec is worse than no spec — it lies. Treat drift as a bug, not a normal outcome.
When To Use
Use this skill when:
- A
specs/<id>/TECH.mdalready exists and is approved (either by the user or by a previouswrite-tech-specrun) - The user says "implement spec X" or "follow the plan in specs/Y/"
- You're picking up an in-progress spec implementation (PROJECT_LOG.md exists)
Do not start implementing without a spec — drop back to
write-tech-spec first, or implement directly if the work is too small
to spec.
Operating Loop
For each numbered step in the spec's Implementation Plan:
- Re-read the spec. It may have been updated since you last looked.
- Read the cited current-state code. TECH.md cites
file:lineanchors; verify they still point where the spec claims. If the code has moved, update the spec citations first, then continue. - Implement the step. Stay scoped to what this step describes — don't drag in unrelated cleanup, refactors, or "while I'm here" improvements. Those belong in their own PRs.
- Run the testing scenarios the spec lists for this step. If the spec doesn't list one, add the smallest test that exercises the new behavior end-to-end.
- Update the spec if reality diverged. Common cases:
- You discovered the proposed approach doesn't work → update Proposed Changes and Implementation Plan, then re-read step 1.
- You found a missed dependency / risk → log it in
DECISIONS.md(create if absent) with a one-line rationale. - You finished early and a planned step is no longer needed → strike it from the plan with a note explaining why.
- Commit. Spec edits and code edits go in the same commit. Commit
messages reference the spec id:
feat(<id>): <step summary>.
PROJECT_LOG.md Discipline
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.
- 1mo ago First seen · 127 lines · 59 tokens per session scan A a474f584b74e
implement-specs is a skill published in the GitHub repository qinghaoliqaq/AutoCoder (12 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 59 tokens to every session and 1,129 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 2026-08-30.
Other skills, from other repositories
meta-pre-commit-quality-gate
Run three quality gates (ruff + mypy + pytest) in parallel over the staged diff, then arbitrate a single BLOCK/APPROVE verdict. Use before committing changes locally when you want a comprehensive pre-commit gate beyond per-file linting — exactly the same gate set CI enforces.
commit
Create git commits with good messages. Use when user says "commit", "create commit", or asks to commit changes.
app_tagline_backfill
Give apps built on this instance a share-card tagline when their app.yaml has none. Runs automatically once after Rome upgrades (AUTO mode, list supplied by the caller) or on request from the guardian ("add taglines to my apps", MANUAL mode, confirm before writing). Only touches editable mode: "source" apps — never…
guardian
Gatekeeping Git/PR by classifying change essence and recommending granularity, naming, and strategy; the ship recipe also executes PR delivery (create, CI watch, merge, cleanup) with Ask First on merge. Use when PR preparation, commit strategy, or gated PR delivery is needed.
worktree
Isolate parallel work in a git worktree of the current repository (worktree tool: status / new / merge / remove). Use when another live session is dirty in the same directory (listagents), when two agents would fight over index.lock, or when starting a parallel task checkout.
commit
Generate git commit messages from staged changes.