agentleFS
Cody-SimsCody-Sims/agent-skills/skills/systematic-debugging/SKILL.md

systematic-debugging

Skill0 starsChanged 4 months agoMITView on GitHub

Diagnoses failures by reproducing exact errors, testing one hypothesis at a time, and fixing the root cause. Use when a bug, failing test/build, crash, exception, wrong output, or intermittent failure needs tracing; not for reviewing a patch or auditing already-green tests.

190 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use chinese-code-review, chinese-commit-conventions and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Leodorareluctant259Skill

    chinese-code-review

    中文代码审查规范——在保持专业严谨的同时,用符合国内团队文化的方式给出有效反馈

    Also in 8 projects that use this

  • Leodorareluctant259Skill

    chinese-commit-conventions

    中文 Git 提交规范 — 适配国内团队的 commit message 规范和 changelog 自动化

    Also in 8 projects that use this

  • obraSkill

    diagnosing-superpowers

    Use when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.

    Also in 20 projects that use this

  • mjunaidcaSkill

    building-chatgpt-apps

    Guides creation of ChatGPT Apps with interactive widgets using OpenAI Apps SDK and MCP servers. Use when building ChatGPT custom apps with visual UI components, embedded widgets, or rich interactive experiences. Covers widget architecture, MCP server setup with FastMCP, response metadata, and Developer Mode configuration. NOT when building standard MCP servers without widgets (use building-mcp-servers skill instead).

    Also in 4 projects that use this

  • mjunaidcaSkill

    building-chat-interfaces

    Build AI chat interfaces with custom backends, authentication, and context injection. Use when integrating chat UI with AI agents, adding auth to chat, injecting user/page context, or implementing httpOnly cookie proxies. Covers ChatKitServer, useChatKit, and MCP auth patterns. NOT when building simple chatbots without persistence or custom agent integration.

    Also in 4 projects that use this

  • +2 more from Leodorareluctant259
  • +5 more from mjunaidca

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 177 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read applicable instructions and capture the full error, stack/cause chain, exit code, and failing assertion verbatim.

    line 22
  2. Distinctive:

    If a probe or fix partially fails or its write outcome is uncertain, read mutation recovery before another mutation.

    line 50
  3. Distinctive:

    If repeated fixes expose unrelated failures, stop patching and return to step 3: recheck the mental model, coupling, and known-good …

    line 56

“Common” is measured: kept by at least 6 in 10 of the 177 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Systematic Debugging

Contract

Inputs: observed versus expected behavior, failure command/input, and relevant environment. Output: a confirmed cause and verified fix with a regression test, or a precise blocker. No speculative production edits before reproduction and an evidence-backed explanation; diagnostic probes are allowed.

Workflow

  1. Read applicable instructions and capture the full error, stack/cause chain, exit code, and failing assertion verbatim. Preserve exact diagnostic evidence without exposing secrets; identify any redactions.
  2. Find and run the smallest deterministic reproduction using existing commands. Record command, input, and environment. For intermittency, read flaky tests; bound attempts/time and capture failing seed/order/state. A missing tool or dependency is a readiness problem, not proof of an application defect: follow documented, authorized setup for demonstrated missing prerequisites, retry the same command, or report blocked.
  3. Narrow the search to the error location, input/output path, recent changes, and differences from a known-good state. Use symbols or path-scoped searches with bounded output; expand only for evidence. Do not launch workers or map unrelated architecture for one continuous failure trace.
  4. State one falsifiable hypothesis with the predicted observation. Test it with the smallest boundary probe, breakpoint, or bisection; confirm or reject it before another hypothesis. Trace bad state backward to its first divergence, not just the final throwing frame. Read root-cause tracing when stack analysis, instrumentation, history/input bisection, or differential diagnosis is needed.
  5. Correct the proven origin. Do not swallow errors, clamp bad state, add blind retries, or special-case the observed input to hide the symptom. Remove diagnostic instrumentation when finished. Add a regression test and show it fails without the fix for the expected reason, then passes with the fix.
  6. Rerun the exact reproduction, then affected surrounding tests and required repository gates. Combine compatible selectors rather than repeating full suites after each probe. For flakes, report bounded runs/seeds and residual intermittency; one green attempt does not prove nondeterminism gone.

If a probe or fix partially fails or its write outcome is uncertain, read mutation recovery before another mutation. Reconcile actual state and rerun the relevant probe.

Stop and output

If repeated fixes expose unrelated failures, stop patching and return to step 3: recheck the mental model, coupling, and known-good differences. If reproduction, prerequisites, or diagnosis remain unavailable, report the blocker rather than claiming a fix.

Return reproduction command/input, exact relevant error, confirmed cause and evidence, fix, regression red/green results, final commands/results, and remaining uncertainty. Do not infer success from inspection or omit an unresolved verdict.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.