用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/ahrav/scratch-scanner-rs --skill run-fuzz命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
Write-then-verify documentation pipeline. Use when a user asks to improve comments or docs, explain algorithms or design choices, write or upgrade docstrings, or raise documentation quality for a codebase (especially Rust crates). Writes docs, then automatically verifies every claim against code reality using a fresh agent to eliminate confirmation bias.
Use when you have code review findings, PR comments, or review reports that need to be systematically addressed — especially when there are multiple findings across different files and severities
Use when creating any beads task — auto-researches the codebase, links related tasks, and produces a rich self-contained description from a structured template. Accepts minimal intent and outputs a complete task ready for agent implementation.
正在显示 SKILL.md
基于 SOC 职业分类
| name | run-fuzz |
| description | Run cargo-fuzz targets with proper nightly toolchain and options |
Run fuzz tests against the scanner's parsing and decoding logic.
run-fuzz <target> - Run specific fuzz target (default 10,000 runs)run-fuzz <target> <runs> - Run with custom iteration countrun-fuzz all - Run all targets briefly (1,000 runs each)run-fuzz list - List available targets| Target | Tests |
|---|---|
fuzz_anchor_soundness | Anchor extraction correctness |
fuzz_b64_gate_build_and_scan | Base64 gate construction and scanning |
fuzz_b64_gate_determinism | Base64 gate deterministic behavior |
fuzz_b64_gate_differential | Base64 gate differential testing |
fuzz_tiger_chunking | Tiger harness chunking logic |
cd fuzz && cargo +nightly fuzz run <target> -- -runs=10000
cd fuzz
for target in fuzz_anchor_soundness fuzz_b64_gate_build_and_scan fuzz_b64_gate_determinism fuzz_b64_gate_differential fuzz_tiger_chunking; do
echo "Running $target..."
cargo +nightly fuzz run $target -- -runs=1000
done
Crashes are saved to fuzz/artifacts/<target>/. To reproduce:
cargo +nightly fuzz run <target> fuzz/artifacts/<target>/<crash-file>
rustup install nightlycargo install cargo-fuzz