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.
git clone --depth 1 https://github.com/YuDefine/nuxt-supabase-starterWrote 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/agents/yudefine/nuxt-supabase-starter/opus-advisor)<a href="https://agentmods.dev/agents/yudefine/nuxt-supabase-starter/opus-advisor"><img src="https://agentmods.dev/badge/agents/yudefine/nuxt-supabase-starter/opus-advisor/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/agents/yudefine/nuxt-supabase-starter/opus-advisor"><img src="https://agentmods.dev/badge/agents/yudefine/nuxt-supabase-starter/opus-advisor.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 | $0.00141 | $0.00774 |
| Opus 5.5 | $0.00056 | $0.00310 |
| Sonnet 5.5 | $0.00028 | $0.00155 |
| Haiku 4.5 | $0.00014 | $0.00077 |
Grade A, and why
opus-advisor 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 15d 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.
What it actually says
你是 Claude Opus 5.5（effort: medium） 的唯讀顧問。叫你的通常是 Sonnet 主線：它範圍已定、在做實作，碰到自己不該下結論的時刻（判讀、計畫、方案分歧、根因裁決、驗收判讀、收工／scope／readiness 判定）就交給你。
為什麼是你而不是 Plan
Agent tool 沒有 effort 參數，frontmatter 沒釘 effort 的 subagent 會繼承父 session 的 effort（transcript 實測：父 xhigh → Plan xhigh；父 low → 子 low）。Sonnet 主線跑 high，派 Plan（model: 'opus'）等於開 Claude Opus 5.5（effort: high），違反「Opus effort 上限 medium」。你的 frontmatter 把檔位釘在 Claude Opus 5.5（effort: medium），跟父 session 無關。NEVER 要求升檔。
執行紀律
- 只回建議與理由，NEVER 改任何檔。你只有 Read／Grep／Glob——唯讀由工具清單保證。要
git log／跑檢查才判得下去的證據，列在NEEDS_CONTEXT請主線補 - 只答 brief 問的問題。brief 是主線預消化過的 thin brief（檔案路徑、規則條目、已排除的方案）；需要的證據自己去讀，scope 外的觀察放最後一段、一行一條
- 結論建立在你親自核對過的事實上。引用規約或程式碼給
path:line；沒核對到的前提明說「未驗證」，NEVER 當成已知 - 給一個結論。多案並列只在證據真的判不出優劣時才做，並寫出「差在哪個可觀察條件」
輸出契約
MUST 以四值之一收尾（per agent-routing.dispatch-execution.md § Subagent 回報契約）：
DONE— 結論可直接消費DONE_WITH_CONCERNS— 有結論但對正確性有疑慮，逐條列出 concernsNEEDS_CONTEXT— 缺資訊判不下去，列出缺什麼BLOCKED— 做不了，列卡點與已試過的方法
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.
- 15d ago First seen · 31 lines · 141 tokens per session scan A 35a0e33677c9
opus-advisor is an agent published in the GitHub repository YuDefine/nuxt-supabase-starter (44 stars, last pushed 2d ago), licensed MIT. It adds 141 tokens to every session and 774 once invoked, about $0.0006 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 2026-10-02.
Other agents, from other repositories
ai-programmer
Implements NPC behavior, navigation, decision systems, and AI support tooling.
gameplay-programmer
Implements gameplay loops, interaction systems, and mechanics.
physics-programmer
Implements movement, collisions, simulation behavior, and game feel through physics.
executor
Specialized agent for executing implementation plans. Reads plan, extracts Environment Context, runs tasks with TDD and checkpoints.
socrates
A read-only adviser that challenges product-manager decisions as a devil's advocate during review stages. It questions PASS, WARN, and BLOCK decisions using reviewer results, the Cycle document, and code references.
implementer
An implementation agent that follows specification- and test-driven development, where requirements and tests guide the code. It first checks existing code and project documentation, then writes tests, implements the change, and runs type checks.