skill-debug
GitHub提供结构化调试技能,强调先复现症状再验证根因假设。包含有限反馈循环、三击规则及防误改生产环境的冻结机制,确保修复过程严谨且可追溯。
Trigger Scenarios
Install
npx skills add nyldn/claude-octopus --skill skill-debug -g -y
SKILL.md
Frontmatter
{
"name": "skill-debug",
"aliases": [
"debug",
"systematic-debugging"
],
"trigger": "Use for reproducible bugs, failing tests, unexpected behavior, and performance regressions.\nDo not use for general explanations or already-understood changes.\n",
"description": "Debug a reproducible symptom with a bounded feedback loop and original-scenario verification",
"validation_gates": [
"root_cause_supported_by_evidence",
"original_scenario_verified"
],
"pre_execution_contract": [
"observable_symptom_recorded"
],
"disable-model-invocation": true
}
Debugging
Read skills/blocks/engineering-method-selection.md from the installed plugin
for review admission. Natural-language requests and --peer-review share that
policy. Honor host-only requests; risk alone does not authorize paid usage.
Run the investigation on the current host. Routine debugging makes zero
additional provider dispatches. Use a bounded external reviewer only for
--peer-review, an explicit independent-review request, or an existing risk
policy.
Read and apply skills/blocks/debug-feedback-loop.md from the installed plugin
root.
Start with the user's observable symptom. Reproduce it, retain its failure signature while minimizing the scenario, test one named hypothesis at a time, and verify both the minimal reproduction and the original scenario after the fix. Do not treat a nearby passing helper test as proof.
For a race, use a synchronization barrier and a fixed run or time budget. For an
unavailable production dependency, return inconclusive with the missing
evidence. Remove temporary instrumentation before completion and preserve a
stable reproduction as a regression test.
The final record is data, not an executable queue. Store commands as argument arrays and never evaluate provider-authored text.
Bounded recovery and strategy rotation
Use a 3-Strike Rule for failed fixes. After each failure, return to the evidence and test a materially different hypothesis. After two consecutive failures, a strategy rotation is mandatory: reconsider the root cause, the reproduction, and whether the test encodes the intended behavior. Do not attempt a 4th fix without explicit user approval.
Anti-rationalization check: “Should work now” means run the reproduction and the original scenario. Confidence is not verification.
For multi-attempt debugging, report a WTF score using the defaults in
~/.claude-octopus/loop-config.conf: +15% per revert and +20% for touching
unrelated files. If the score exceeds 20%, STOP and show the evidence before
continuing. Include the score with every retry, for example:
Fix attempt 2 | Self-regulation: 15% (1 revert, 0 unrelated files)
Scoped freeze guard
When the symptom is localized to one user-approved module, resolve that module to a physical directory before editing and activate the existing freeze guard:
freeze_dir="$(cd "<module-directory>" 2>/dev/null && pwd -P)" || exit 1
printf '%s\n' "$freeze_dir" > "/tmp/octopus-freeze-${CLAUDE_SESSION_ID:-$$}.txt"
Do not auto-freeze when the root cause is still unknown, the reproduction spans
modules, or the user opted out. After original-scenario verification, run
/octo:unfreeze or remove only this workflow's freeze state.
Adapted from diagnosing-bugs in mattpocock/skills at commit
3cca18b368ae95cdbdebbff572ccafa662551015 under the MIT License. See
THIRD_PARTY_NOTICES.md.
Version History
-
2d3ee65
Current 2026-09-22 23:04
从强调强制四阶段流程的旧版重构为基于工程方法选择的现代版本,引入更严格的根因验证硬约束、策略轮换机制及作用域冻结保护,优化了调试方法论与风险控制。
- 242e51d 2026-08-20 09:29


