testing-handbook-skills: Скилл для тестирования безопасности приложений

--- name: testing-handbook-skills description: "Application security testing toolkit from the Trail of Bits Testing Handbook. Helps the agent set up fuzzing campaigns, write fuzz harnesses, run coverage-guided fuzzers (libFuzzer, AFL++, cargo-fuzz, Atheris, Ruzzy), and triage crashes. Covers memory-safety sanitizers (AddressSanitizer, UBSan, MSan), static analysis with Semgrep and CodeQL, cryptographic validation using Wycheproof test vectors, and constant-time verification. Use when testing C, C++, Rust, Python, or Ruby code for vulnerabilities, improving code coverage, building seed corpora, creating fuzzing dictionaries, overcoming fuzzing obstacles, or integrating security checks into CI/CD with OSS-Fuzz." --- # Testing Handbook Skills Comprehensive security testing toolkit generated from the [Trail of Bits Application Security Testing Handbook](https://appsec.guide/). ## When to Use - Setting up fuzzing campaigns for C/C++, Rust, Python, or Ruby - Writing fuzzing harnesses for target functions - Analyzing code coverage to guide testing - Running sanitizers (AddressSanitizer, UBSan, MSan) to catch memory bugs - Performing constant-time testing for cryptographic code - Using Wycheproof test vectors for crypto validation ## When NOT to Use - Smart contract auditing or chain-specific review work - Writing custom static-analysis rules from scratch - General code review outside a security-testing workflow - Vulnerability hunting without a concrete testing plan ## Sub-Skills (17 total) ### Fuzzers | Fuzzer | Language | Best For | Skill Path | |--------|----------|----------|------------| | **libFuzzer** | C/C++ | LLVM-based coverage-guided fuzzing | [skills/libfuzzer/SKILL.md](skills/libfuzzer/SKILL.md) | | **AFL++** | C/C++ | Advanced mutation-based fuzzing | [skills/aflpp/SKILL.md](skills/aflpp/SKILL.md) | | **libAFL** | C/C++ | LibAFL-based custom fuzzers | [skills/libafl/SKILL.md](skills/libafl/SKILL.md) | | **cargo-fuzz** | Rust | Rust native fuzzing with libFuzzer backend | [skills/cargo-fuzz/SKILL.md](skills/cargo-fuzz/SKILL.md) | | **Atheris** | Python | Python coverage-guided fuzzing | [skills/atheris/SKILL.md](skills/atheris/SKILL.md) | | **Ruzzy** | Ruby | Ruby coverage-guided fuzzing | [skills/ruzzy/SKILL.md](skills/ruzzy/SKILL.md) | ### Techniques | Technique | Purpose | Skill Path | |-----------|---------|------------| | **Harness Writing** | Writing effective fuzzing harnesses | [skills/harness-writing/SKILL.md](skills/harness-writing/SKILL.md) | | **Coverage Analysis** | Measuring and improving code coverage | [skills/coverage-analysis/SKILL.md](skills/coverage-analysis/SKILL.md) | | **Fuzzing Dictionary** | Creating effective fuzzing dictionaries | [skills/fuzzing-dictionary/SKILL.md](skills/fuzzing-dictionary/SKILL.md) | | **Fuzzing Obstacles** | Overcoming common fuzzing barriers | [skills/fuzzing-obstacles/SKILL.md](skills/fuzzing-obstacles/SKILL.md) | | **AddressSanitizer** | Memory error detection with ASan | [skills/address-sanitizer/SKILL.md](skills/address-sanitizer/SKILL.md) | ### Static Analysis | Tool | Purpose | Skill Path | |------|---------|------------| | **Semgrep** | Fast pattern-matching security scans | [skills/semgrep/SKILL.md](skills/semgrep/SKILL.md) | | **CodeQL** | Deep semantic code analysis | [skills/codeql/SKILL.md](skills/codeql/SKILL.md) | ### Cryptographic Testing | Tool | Purpose | Skill Path | |------|---------|------------| | **Wycheproof** | Test vectors for crypto implementations | [skills/wycheproof/SKILL.md](skills/wycheproof/SKILL.md) | | **Constant-Time Testing** | Verify constant-time crypto properties | [skills/constant-time-testing/SKILL.md](skills/constant-time-testing/SKILL.md) | ### Infrastructure | Tool | Purpose | Skill Path | |------|---------|------------| | **OSS-Fuzz** | Google's continuous fuzzing service | [skills/ossfuzz/SKILL.md](skills/ossfuzz/SKILL.md) | ### Meta | Tool | Purpose | Skill Path | |------|---------|------------| | **Generator** | Generate new skills from the Testing Handbook | [skills/testing-handbook-generator/SKILL.md](skills/testing-handbook-generator/SKILL.md) | ## Workflow ### Starting a fuzzing campaign 1. **Choose a fuzzer** based on your target language (see Fuzzers table) 2. **Write a harness** using the harness-writing skill 3. **Build with sanitizers** (AddressSanitizer recommended as baseline) 4. **Create a seed corpus** with representative inputs 5. **Run the campaign** and monitor coverage 6. **Analyze coverage** to find uncovered code and improve the harness 7. **Triage crashes** and deduplicate findings ### Setting up CI/CD testing 1. **OSS-Fuzz** for open-source projects (continuous fuzzing) 2. **Semgrep + CodeQL** for static analysis in PRs 3. **Wycheproof** test vectors for crypto validation ## Quick Start by Language | Language | Fuzzer | Harness | Sanitizer | |----------|--------|---------|-----------| | C/C++ | libFuzzer or AFL++ | `LLVMFuzzerTestOneInput` | ASan + UBSan | | Rust | cargo-fuzz | `fuzz_target!` macro | Built-in sanitizers | | Python | Atheris | `atheris.FuzzedDataProvider` | N/A | | Ruby | Ruzzy | `ruzzy` harness pattern | N/A | ## Source Material Generated from the [Trail of Bits Application Security Testing Handbook](https://appsec.guide/) using the testing-handbook-generator meta-skill.