Przejdź do treści
Biblioteka skilli
Programowanie · Analiza

advanced-debugging

/advanced-debugging

Diagnozuje złożone błędy: wyjątki, wyścigi, wycieki pamięci i błędy czasowe, na podstawie dowodów z działania programu.

Licencja
MIT
Ostatnia zmiana repozytorium
13.11.2026

Ocena i bezpieczeństwo

90/ 100Klasa: dobry
Format25 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść25 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
2
Skrypty
0
Komentarze HTML
0
Adresy URL w SKILL.md
0
Przestarzałość

Brak sygnałów. Repozytorium zmienione 13.11.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co robi i kiedy

Diagnozuje złożone błędy: wyjątki, wyścigi, wycieki pamięci i błędy czasowe, na podstawie dowodów z działania programu.

Oryginalny opis (EN)

Use when diagnosing complex application defects, unhandled exceptions, race conditions, memory leaks, timing bugs, or unexpected runtime failures requiring evidence-based root-cause analysis.

Jak działa

  1. Purpose
  2. Activation criteria
  3. Non-activation guidelines
  4. Required context and inputs
  5. Operating procedure
  6. Technical and engineering standards
  7. Verification and evidence requirements
  8. Expected outputs and deliverables
  9. Risks, boundaries and failure modes
  10. Progressive disclosure references
Sekcje z nagłówków SKILL.md, w oryginale · 85 linii, 785 słów
Podgląd SKILL.md
---
name: "advanced-debugging"
description: "Use when diagnosing complex application defects, unhandled exceptions, race conditions, memory leaks, timing bugs, or unexpected runtime failures requiring evidence-based root-cause analysis."
---

# Advanced Debugging — Systematic Failure Isolation and Root-Cause Diagnosis

> Generated from `specs/skills/advanced-debugging.md` by `scripts/generate-antigravity-skills.ts`. Edit the canonical specification, not this file.

## Purpose

The Advanced Debugging skill provides a rigorous, hypothesis-driven methodology for diagnosing and resolving complex, intermittent, or deep-seated software defects. It halts speculative trial-and-error editing and enforces evidence-based root-cause isolation, minimal interventions, and compulsory regression test coverage.

## Activation criteria

Activate this skill when:

- An application crashes, throws unhandled exceptions, or enters an error state.
- Tests fail intermittently due to timing, concurrency, or environmental variance.
- Data corruption, race conditions, or state synchronization bugs are observed.
- Performance degrades abruptly or memory leaks are detected.
- External API integrations or persistence queries fail unexpectedly.

## Non-activation guidelines

Do not activate this skill when:

- The compiler or type checker points directly to a missing variable or syntax error.
- The task is building a new feature or extending existing functionality according to a new specification.
- Changing CSS colors, typography, or spacing without underlying logic errors.
- Performing routine dependency upgrades that pass all existing test suites.

## Required context and inputs

Before beginning root-cause isolation, collect:

1. **Observed Behavior:** Exact error message, stack trace, HTTP status code, or visual anomaly.
2. **Expected Behavior:** Clear definition of what the system should have done under these inputs.
3. **Reproduction Prerequisites:** Environment parameters, operating system, runtime version, seed data, and configuration flags.
4. **Execution Evidence:** Console logs, network payloads, server-side log files, and database state snapshots.

## Operating procedure

Execute the following eleven-step debugging workflow:

1. **Understand Observed Behavior:** Record the exact symptoms, error messages, and failure conditions.
2. **Establish Expected Behavior:** Specify the correct invariants and expected outcomes.
3. **Reproduce the Issue:** Create a deterministic, minimal reproduction in a script, unit test, or automated browser test. If non-deterministic, establish conditions that increase failure probability.
4. **Gather Logs, Traces, and Evidence:** Collect runtime stack traces, query logs, variable snapshots, and execution timestamps.
5. **Identify Possible Causes:** Map the error back through the call stack to potential points of invalid state, timing discrepancies, or unhandled nullability.
6. **Form Testable Hypotheses:** State clear, falsifiable hypotheses (_"Component X receives undefined input because Provider Y resolves after Consumer Z renders"_).
7. **Isolate the Root Cause:** Devise focused experiments to test each hypothesis. Discard disproven theories; confirm the true root cause with verifiable evidence.
8. **Implement the Smallest Appropriate Fix:** Change only the code directly responsible for the defect. Avoid shotgun refactoring or touching unrelated files.
9. **Add a Regression Test:** Author an automated test that reproduces the defect and fails without the fix, then passes with it.
10. **Run Relevant Verification:** Execute the entire affected test suite and related integration tests to confirm no regressions were introduced.
11. **Report Remaining Uncertainty:** Clearly disclose if any edge conditions remain unverified or if underlying environment factors could not be reproduced locally.

## Technical and engineering standards

- **Evidence First:** Never claim a defect is resolved because a proposed edit "looks correct." Execution evidence (test output, log confirmation) is mandatory.
- **Minimal Surface Area:** A bug fix should touch the minimum number of lines and files required. Do not bundle opportunistic refactorings into a defect patch.
- **Idempotent Resource Cleanup:** When debugging resource exhaustion, ensure all file descriptors, network sockets, timers, and database connections are closed in `finally` blocks or structured teardowns.
- **Defensive Invariant Checks:** Validate assumptions at component boundaries with explicit assertions or schema validation rather than swallowing unexpected values silently.

## Verification and evidence requirements

- Provide the exact reproduction test command and terminal output showing failure before the fix.
- Provide the test command and terminal output showing clean pass after the fix.
- Inspect `git diff` to verify that zero unintended edits or extraneous whitespace modifications were made.

## Expected outputs and deliverables

- **Root Cause Analysis (RCA) Statement:** Clear explanation of why the failure occurred, what assumption was violated, and why the fix works.
- **Focused Code Patch:** Minimal, clean source modifications addressing the defect.
- **Automated Regression Test:** A permanent, reproducible test case checked into the repository.
- **Verification Log:** Terminal output proving test pass and absence of collateral damage.

## Risks, boundaries and failure modes

- **Speculative Editing:** Modifying multiple files simultaneously without proving the root cause obscures the bug and introduces new defects.
- **Masking Symptoms:** Catching and silencing exceptions (e.g., empty `catch {}` blocks or default fallback values) hides bugs rather than solving them.
- **Test Weakening:** Modifying test assertions or removing assertions to make a test suite pass is an automatic failure.

## Progressive disclosure references

For diagnostic frameworks for async race conditions, memory leaks, and connection starvation:

- [Advanced Debugging Playbook: Root-Cause Analysis and Hypotheses Testing](./references/root-cause-analysis-guide.md)
Pliki w paczce (2)
  • SKILL.md 6 KB
  • references/root-cause-analysis-guide.md 3 KB

Wersje

W bazie nie ma innych skilli o nazwie advanced-debugging.

Migawka bazy z 13.11.2026 · wersja paczki e3a792dad45b (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 0 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 100root-cause-debugging
    Diagnozuje incydenty produkcyjne, powtarzające się błędy i wyścigi na podstawie dowodów, a poprawki usuwają mechanizm przyczyny.
  • 91memory-timing-debugging
    Diagnozuje błędy pamięci i czasu w firmware: wycieki, drgania czasowe, głodzenie watchdoga, inwersję priorytetów i zakleszczenia.
  • 90dev-debug
    Systematycznie diagnozuje awarie oprogramowania na podstawie dowodów: testy, błędy runtime, buildy, logi i zrzuty ekranu.
  • 100dart-fix-runtime-errors
    Diagnozuje błędy wykonania w Dart: rzutowania typów, naruszenia sprawdzeń null i nieobsłużone wyjątki asynchroniczne.
  • 100vscode-extension-debugger
    Debuguje i naprawia błędy w rozszerzeniach VS Code: wycieki pamięci, webview, aktywację i komunikację.
  • 100clay-advanced-troubleshooting
    Stosuje pogłębione techniki debugowania Clay przy trudnych do zdiagnozowania problemach: złożone wyścigi i pakiety dowodów do eskalacji do wsparcia Clay.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.