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 AbyssCN/oh-my-dag --skill investigategit clone --depth 1 https://github.com/AbyssCN/oh-my-dagWrote 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/abysscn/oh-my-dag/investigate)<a href="https://agentmods.dev/skills/abysscn/oh-my-dag/investigate"><img src="https://agentmods.dev/badge/skills/abysscn/oh-my-dag/investigate/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/abysscn/oh-my-dag/investigate"><img src="https://agentmods.dev/badge/skills/abysscn/oh-my-dag/investigate.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 | $0.00077 | $0.00987 |
| Opus 5.5 | $0.00031 | $0.00395 |
| Sonnet 5.5 | $0.00015 | $0.00197 |
| Haiku 4.5 | $0.00008 | $0.00099 |
Grade A, and why
investigate 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 4d 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
/investigate — 系统化根因调试
铁律:无根因不修。 修症状 = 打地鼠。找到根因再动手,诊断阶段禁提修复方案。
纯方法论 skill(不封特定 MCP 工具,用 runtime 原生 Grep/Read/git;需并行跑多假设时可派多个子 agent)。
Phase 1 — 历史搜索(先查后调)
同一消息并行发:git log --oneline -20 -- <相关文件>(近期变更)+ Grep 既有 docs/plan/·NOTES·error 记录搜错误关键词(有先例走 /omd-recall 先召回)。命中已知修复 → 直接引用,跳到修复。
Phase 2 — 复现
确定性触发 + 存证据(错误信息/堆栈/复现步骤)。信息不足 → 逐个追问(一次一问)。不能稳定复现 → 收集更多证据再继续,别猜。
Phase 3 — Scope Lock
从症状追到最窄受影响目录,声明范围;此后 Edit 只在范围内,越界先问 owner。危险信号立即停:「顺手修个相关的」(不顺手,记 follow-up)、「既然在改不如重构」(不重构,本 bug 修完再评估)。
Phase 4 — 模式匹配
先 step-back 自问:这类症状(数据不一致/超时/类型错/竞态/权限泄漏/schema 漂移)在本技术栈通常由哪层引起?建宏观诊断框架再缩范围,别直接跳到最"像"的模式。同文件反复修 bug = 架构问题,不是巧合。
Phase 5 — 假设验证(最多 3 个)
每假设三步推导:观察(引 file:line 我看到什么)→ 推论(若成立还应看到什么)→ 验证(设计能区分成立/不成立的实验)。追踪表标 CONFIRMED/REJECTED。先验证再修(可疑根因处加临时 log/断言,跑复现看证据是否吻合)。三振出局:3 个假设全败 → STOP 问 owner(A 继续有新假设 / B 升级人工 / C 埋点观察)。危险信号:还没追数据流就提修复 = 在猜。
Phase 6 — 修复
根因确认后最小改动:最少文件、最少行,不顺手重构相邻代码。写回归测试(不修 FAIL、修后 PASS),跑测试套件贴输出。改 >5 文件 → 问 owner(bugfix 爆炸半径偏大:继续 / 拆分 / 重想)。
Phase 7 — 验证闭环
复现原始场景确认已修 + 跑变更相关验证。不说"应该能修好",验证并证明。
Phase 8 — 报告 + 扫同类
报告:症状 / 根因(为什么)/ 模式 / 修复(file:line)/ 证据 / 回归测试 / 状态(DONE · DONE_WITH_CONCERNS · BLOCKED)。修完扫一片:用根因反推同类——Grep 同 pattern(根因是"缺 scope 过滤" → 全仓找所有 db 访问点;"未校验入参" → 找所有 untrusted 入口)。>5 处别一次全修,修 1-2 关键,其余记 NOTES 留 follow-up。
与既有 skill 的边界
- investigate = 某失败的根因调查。审一批 diff 找 bug → /omd-review;扫安全漏洞 → /security-review;找过度工程 → /ponytail-review。修复方案涉及大改前可用 /grill 审。
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.
- 4d ago First seen · 38 lines · 77 tokens per session scan A 5f7ec2bdc87e
investigate is a skill published in the GitHub repository AbyssCN/oh-my-dag (43 stars, last pushed 4d ago), licensed MIT. It adds 77 tokens to every session and 987 once invoked, about $0.0003 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 2026-10-12.
Other skills, from other repositories
neuron-test
Write tests for Neuron AI agents, RAG systems, workflows, and tools using the built-in testing utilities. Use this skill when the user mentions testing agents, writing unit tests, mocking AI providers, testing tool execution, verifying RAG retrieval, testing workflow behavior, or creating test cases for Neuron AI…
test-hang-watchdog
Prevent and diagnose hanging or long-running Gradle unit tests (> 30s) in Android/Kotlin coroutines codebases. Use when running tests, when a test execution takes unexpectedly long, when background test tasks appear frozen, or when taking thread dumps (jstack) to unblock coroutine deadlocks.
prevent-recurrence
Use when correcting an issue or failure reveals a reusable prevention lesson, including code, setup, measurement, validation, workflow, agent mistakes, or accepted code-review findings; when someone says "this keeps happening" or asks how to stop it; or in post-incident follow-up. Fixes the instance (failing-first…
bounded-code-repair
Repair one focused, inspectable code behavior with a declared file boundary, actual validation, and a bounded retry count. Do not use for broad feature delivery, architecture decisions, security review, or documentation-only work.
fix-bug
Diagnoses and fixes bugs using a test-first workflow (reproduce, diagnose, fix). Use this skill whenever the user reports a bug, describes unexpected behavior, or asks to investigate or fix a defect. Even if the user says "something's broken", "this isn't working", "fix this bug", or "why does X happen", load this…
dart-flutter-equatable-engineering
Design, implement, review, debug, test, migrate, and optimize value equality in production Dart and Flutter code using Equatable or appropriate alternatives. Use this skill whenever work involves Equatable, props, immutable Bloc/Cubit states or events, equality bugs, suppressed state emissions, lists/sets/maps in…