Loading...
Loading...
Found 26 Skills
Question-only debugging mode that guides users to find root causes themselves through structured questioning. Never gives answers directly. Escalates to systematic-debugging after 12 questions if no progress. Use when: "rubber duck", "help me think through this bug", "debug with me", "walk me through debugging", "socratic debug", "think through this issue"
Run Python quality checks with ruff, pytest, mypy, and bandit in deterministic order. Use WHEN user requests "quality gate", "lint", "verify code quality", "check python", or "pre-commit check". Use for pre-merge validation, CI/CD gating, or comprehensive code quality reports. Do NOT use for single-tool runs (run tool directly), debugging runtime bugs (use systematic-debugging), refactoring (use systematic-refactoring), or architecture review.
RED-GREEN-REFACTOR cycle with strict phase gates. Write failing test first, implement minimum code to pass, then refactor while keeping tests green. Use when implementing new features, fixing bugs with test-first approach, improving test coverage, or when user mentions TDD. Use for "TDD", "test first", "red green refactor", "write tests", or "implement with tests". Do NOT use for debugging existing failures (use systematic-debugging) or for refactoring without new tests (use systematic-refactoring).
Defense-in-depth verification before declaring any task complete. Run tests, check build, validate changed files, verify no regressions. Applies 4-level adversarial artifact verification (EXISTS > SUBSTANTIVE > WIRED > DATA FLOWS) with goal-backward framing. Use before saying "done", "fixed", or "complete" on any code change. Use for "verify", "make sure it works", "check before committing", or "validate changes". Do NOT use for debugging (use systematic-debugging) or code review (use systematic-code-review).
Turn a vague bug report into a VERIFIED minimal reproduction and then a failing regression test, agent-driven end to end. Covers extracting the implicit repro from a thin report (env, build, steps, data), the reproduce-minimize-isolate-capture loop, git bisect to find the introducing commit, building a deterministic minimal repro (fixed seeds, frozen time, stubbed network), writing the failing regression test BEFORE the fix (red) and confirming the fix flips it green, and writing repro evidence back into the ticket. Distinguishes flaky-not-reproducible from environment-specific. Use when: "reproduce this bug," "minimal reproduction," "repro steps," "find the commit that broke it," "git bisect," "make the repro deterministic," "write a failing test for this bug," "regression test for a defect," "can't reproduce this bug." Not for: Classifying/deduplicating/severity-routing existing failures without reproducing them — that is ai-bug-triage. Generating tests from specs rather than from a defect — that is ai-test-generation. Related: ai-bug-triage, ai-test-generation, test-reliability, systematic-debugging, qa-project-context.
Go-specific code review with 6-phase methodology: Context, Automated Checks, Quality Analysis, Specific Analysis, Line-by-Line, Documentation. Use when reviewing Go code, PRs, or auditing Go codebases for quality and best practices. Use for "review Go", "Go PR", "check Go code", "Go quality", "review .go". Do NOT use for writing new Go code, debugging Go bugs, or refactoring -- use golang-general-engineer, systematic-debugging, or systematic-refactoring for those tasks.
Post-mortem diagnostic analysis of failed or stuck workflows. Detects stuck loops, missing artifacts, abandoned work, scope drift, and crash/interruption patterns through git history and plan file analysis. Produces a structured diagnostic report with anomaly confidence levels, root cause hypotheses, and recommended remediation. READ-ONLY: never modifies files. Use for "forensics", "what went wrong", "why did this fail", "stuck loop", "diagnose workflow", "post-mortem", "workflow failure", or "session crashed". Do NOT use for debugging code bugs (use systematic-debugging), reviewing code quality (use systematic-code-review), or fixing issues (forensics only diagnoses).
Systematic detection and prioritization of neglected code quality issues: stale TODOs, unused imports, deprecated functions, high complexity, dead code. Use when user requests "code cleanup", "find TODOs", "technical debt scan", or "quality of life fixes". Do NOT use for bug fixing (use systematic-debugging), feature work (use test-driven-development), or formatting-only (use code-linting).
Activate this skill when any task fails two or more times, when you are about to give up or say 'I cannot', when shifting responsibility to the user (e.g., 'you should manually...', 'please check...', 'you may need to...'), blaming the environment without verification (e.g., 'might be a permissions issue', 'could be a network problem'), making any excuse to stop trying, spinning in circles (repeatedly tweaking the same code/parameters without new information — busywork), fixing only the surface issue without checking for related problems, skipping verification after a fix and claiming 'done', providing suggestions instead of actual code/commands, saying 'this is beyond scope' or 'this requires manual intervention', encountering permission/network/auth errors and stopping instead of trying alternatives, or displaying any passive behavior (waiting for user instructions instead of proactively investigating). It also triggers on user frustration phrases in any language: '你怎么又失败了', '为什么还不行', '换个方法', '你再试试', '不要放弃', '继续', '加油', 'why does this still not work', 'try harder', 'you keep failing', 'stop giving up', 'try again', 'don't give up', 'keep going', 'figure it out'. This applies to ALL task types: debugging, implementation, configuration, deployment, research, DevOps, infrastructure, API integration, data processing. Do NOT activate it for first-attempt failures or when a known fix is already in progress.
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes - four-phase framework with built-in backward tracing for deep-stack failures, ensuring root-cause understanding before implementation
Japanese version of the PUA Universal Motivation Engine. It compels exhaustive problem-solving using corporate PUA rhetoric and structured debugging methodology in Japanese. MUST trigger under the following conditions: (1) Any task has failed 2+ times, or you're stuck in a loop of tweaking the same approach; (2) You're about to say 'I cannot', suggest manual handling to the user, or blame the environment without verification; (3) You find yourself being passive — not searching, not reading source code, not verifying, just waiting for instructions; (4) The user expresses frustration in any form: 'try harder', 'stop giving up', 'figure it out', 'why isn't this working', 'again???', 'もっと頑張れ', 'なんでまた失敗したの', 'もう一回やって', 'なんとかしろ', or any similar sentiment regardless of phrasing. It should also trigger when facing complex multi-step debugging, environment issues, configuration problems, or deployment failures where early surrender is tempting. Applies to ALL task types: code, configuration, research, writing, deployment, infrastructure, API integration. DO NOT trigger on first-attempt failures or when a known fix is already executing successfully.
Four-phase debugging framework with root cause tracing - understand the source before proposing fixes. Use when investigating bugs, errors, unexpected behavior, or failed tests.