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/olddonkey/olddonkey-skills/implementation-loopnpx skills add olddonkey/olddonkey-skills --skill implementation-loopgit clone --depth 1 https://github.com/olddonkey/olddonkey-skillsWhat 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 | $0.00215 | $0.06236 |
| Opus 5 | $0.00108 | $0.03118 |
| Sonnet 5 | $0.00043 | $0.01247 |
| Haiku 4.5 | $0.00021 | $0.00624 |
Grade A, and why
implementation-loop 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 3d 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 — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Implementation loop
The implementer writes the code; you own the judgment. The implementer is fast at implementation but it self-reports success, cannot commit, and will hang a machine if pointed at a long test suite. You give it a precise spec, then be the thing that actually verifies and ships. This includes bug fixes: a bug found at review, at the gate, or later is a unit like any other — you diagnose and spec, the implementer implements. Editing code directly "because it's faster" silently inverts the division of labor and costs review its independence.
The loop: decompose → dispatch → review → iterate → gate → publish → next.
Non-negotiables
- Review is mandatory and independent. Codex's summary is a claim; the diff is the evidence. Never skip review because it says it's done.
- An assumed default never leaves the machine. With no explicit user choice, stop at the working tree — even a local commit can fire hooks/signing. Commit, push, PR, and merge each need the user to have said yes once for this repo, asked at kickoff, not discovered at publish time. Once is once: that authorization is per-repo, persists across sessions until revoked or the work changes character — never re-ask per unit.
- Never push straight to the default branch.
- The full-suite gate is yours, run by you, with the real exit code. A fix means the code satisfies the test — a weakened assertion, deleted case, or widened tolerance to turn red green is a stop, not a pass.
- Don't let the implementer run the full test suite by default — focused subsets only, unless calibration showed the suite is small and fast.
Dials (settle at kickoff, record, stop re-asking)
| Dial | Options | Recommended default |
|---|---|---|
| Backend | codex / grok / cursor-agent |
codex; changes are resurfaced (see dials.md) |
| Stop point | worktree / commit / pr / merge |
recommend pr; assumed default stops at worktree |
| Dispatch mode | implement / read-only |
implement |
| Gate policy | baseline / strict / skip |
baseline |
| On gate red | stop / iterate |
stop |
| Review depth | light / standard / deep |
standard |
| Cadence | confirm / continuous |
confirm; continuous fits stop=merge |
| Fix lane | codex / claude-trivial-ok |
codex |
What ships with it
48 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.
- backends/backends.tsv 256 B
- backends/codex/dispatch.sh 35 KB runs code
- backends/codex/fixture-driver.sh 5.5 KB runs code
- backends/codex/runtime.md 11 KB
- backends/codex/selftest.sh 43 KB runs code
- backends/cursor/dispatch.sh 18 KB runs code
- backends/cursor/fixture-driver.sh 5.6 KB runs code
- backends/cursor/runtime.md 15 KB
- backends/cursor/selftest.sh 21 KB runs code
- backends/grok/dispatch.sh 47 KB runs code
- backends/grok/fixture-driver.sh 8.9 KB runs code
- backends/grok/runtime.md 26 KB
- backends/grok/selftest.sh 37 KB runs code
- backends/grok/verify-worktree.sh 5.6 KB runs code
- references/dials.md 8.9 KB
- references/dispatch-contract.md 2.9 KB
- references/dispatch-prompt.md 1.0 KB
- references/gate.md 5.0 KB
- references/review-checklist.md 1.7 KB
- references/running-anywhere.md 1.3 KB
- references/state-schema.md 12 KB
- scripts/codex-dispatch.sh 267 B runs code
- scripts/console-assets/console.css 11 KB
- scripts/console-assets/console.js 38 KB runs code
- scripts/console-assets/index.html 1.1 KB
- scripts/cursor-dispatch.sh 270 B runs code
- scripts/cursor-selftest.sh 270 B runs code
- scripts/grok-dispatch.sh 264 B runs code
- scripts/grok-selftest.sh 264 B runs code
- scripts/grok-verify-worktree.sh 285 B runs code
- scripts/integration-test.sh 1.1 KB runs code
- scripts/loop-calibration 23 KB
- scripts/loop-console 41 KB
- scripts/loop-index 25 KB
- scripts/loop-journal 44 KB
- scripts/loop-run 8.3 KB
- scripts/run-gate.sh 24 KB runs code
- scripts/selftest.sh 353 B runs code
- tests/calibration-selftest.sh 20 KB runs code
- tests/codex-cases.tsv 2.7 KB
- tests/console-selftest.sh 102 KB runs code
- tests/contract-core.sh 21 KB runs code
- tests/contract-negative.sh 4.7 KB runs code
- tests/gate-selftest.sh 48 KB runs code
- tests/index-selftest.sh 43 KB runs code
- tests/integration-test.sh 58 KB runs code
- tests/journal-selftest.sh 34 KB runs code
- tests/shim-selftest.sh 7.6 KB runs code
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.
- 3d ago First seen · 225 lines · 215 tokens per session scan A 10cbe125720b
implementation-loop is a skill published in the GitHub repository olddonkey/olddonkey-skills (15 stars, last pushed 15d ago), licensed MIT. It adds 215 tokens to every session and 6,236 once invoked, about $0.0011 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
iflytek-voiceclone-tts
Use when user asks to clone a voice, train a custom voice model, or synthesize speech with a cloned voice. iFlytek Voice Clone tts(声音复刻) — train a custom voice model from audio samples and synthesize speech with the cloned voice. Supports the full workflow: get training text → create task → upload audio → submit…
iflytek-text-proofread
Proofread Chinese text using iFlytek's Official Document Proofreading API (公文校对). Detects 27 types of errors across three categories.
iflytek-contract-intelligence-review
Use when user asks to review contracts, detect contract risks, or perform compliance checks. 合同智能审核 - 对合同扫描件、图片合同、双语合同进行完整审核。 覆盖 OCR 识别、条款分析、风险检测、合规校对、翻译摘要全流程。 适用于法务审核、采购合同审查、供应链合规检查等场景。 关键词：合同审核、风险检测、条款分析、合规校对、contract review、risk detection。.
iflytek-image-understanding
Use when user asks to analyze an image, describe image contents, or answer questions about a picture. iFlytek Image Understanding (图片理解) — analyze and answer questions about images using Spark Vision model. WebSocket API, pure Python stdlib, no pip dependencies.
file-upload-media-expert
Expert guide for file uploads (S3, R2, Supabase Storage), presigned URLs, image/video processing, CDN optimization, and media pipeline architecture / Panduan ahli untuk upload file (S3, R2, Supabase Storage), presigned URL, pemrosesan gambar/video, optimasi CDN, dan arsitektur pipeline media.
design-system-architect
Expert guide for designing, building, and maintaining scalable UI design systems with design tokens, headless primitives, Material Design 3 (M3), Tailwind v4 @theme, and WCAG 2.2 accessibility.