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/Gossip-rs --skill security-reviewer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。
正在显示 SKILL.md
基于 SOC 职业分类
| name | security-reviewer |
| description | Audit memory safety and security in unsafe code blocks, buffer handling, and security-sensitive operations |
Review unsafe code and security-sensitive operations in the gossip-rs codebase.
unsafe blockcrates/gossip-stdx/src/inline_vec.rs - Stack-backed collection with unsafe pointer opscrates/gossip-stdx/src/ring_buffer.rs - Fixed-capacity circular queue with unsafecrates/gossip-stdx/src/byte_slab.rs - Pre-allocated byte pool with raw pointerscrates/gossip-connectors/src/filesystem.rs - Filesystem I/O with unsafecrates/gossip-coordination/src/lib.rs - Coordination protocol internalscrates/scanner-engine/src/engine/hit_pool.rs - Sorted hit pool with unsafe pointer opscrates/scanner-engine/src/engine/scratch.rs - Reusable scratch memory with raw pointerscrates/scanner-engine/src/engine/stream_decode.rs - Streaming decoder with buffer manipulationcrates/scanner-engine/src/engine/buffer_scan.rs - Buffer scanning with unsafe slice opscrates/scanner-engine/src/engine/transform.rs - Transform/decode pipeline (base64, etc.)crates/scanner-engine/src/engine/vectorscan_prefilter.rs - FFI boundary with vectorscancrates/scanner-engine/src/scratch_memory.rs - Scratch memory allocation with raw pointerscrates/scanner-engine/src/lsm/set_associative_cache.rs - Cache with SIMD tag matchingcrates/scanner-engine/src/engine/simd_classify.rs - SIMD byte classificationcrates/scanner-scheduler/src/runtime.rs - Runtime with async I/O and unsafecrates/scanner-scheduler/src/scheduler/ - Work-stealing scheduler (48 files)crates/scanner-git/src/ - Git pack parsing, delta decoding (86 files with binary data handling)get_uncheckedFor each unsafe block:
## Security Review: [file/module]
### Unsafe Block Audit
| Location | Purpose | Safety Justification | Status |
|----------|---------|---------------------|--------|
| line:XX | ptr arithmetic | bounds checked at line:YY | SAFE |
| line:XX | transmute | MISSING JUSTIFICATION | REVIEW |
### Findings
| Severity | Issue | Location | CWE |
|----------|-------|----------|-----|
| CRITICAL | Unchecked bounds | line:XX | CWE-125 |
| HIGH | Integer overflow | line:XX | CWE-190 |
### Recommendations
1. **[Issue]**: Add bounds check before unsafe access
```rust
// Before (unsafe)
// After (safe)
## Related Resources
- `docs/kani-verification.md` - Existing Kani proofs
- `/test-strategy` - Choose appropriate verification approach
- `/run-fuzz` - Run fuzz targets for security testing
Consolidate all /review-capture drop files for the current branch into a single verified, deduplicated, conflict-annotated merged plan. Verifies each finding against HEAD (discarding stale ones), merges duplicates across reviewers, flags conflicting suggested fixes, groups findings into execution waves by file ownership, and deletes the individual drop files on success. Run after all parallel review terminals have completed.
Wrapper skill that invokes any review skill/command and captures its findings into a structured YAML drop file under .claude/review-drops/<branch>/. Use when running parallel code reviews across multiple terminals (each terminal captures one reviewer's output) so a later /merge-reviews pass can dedup, verify, and consolidate before execution. Accepts any target review skill (e.g. ce:review, multi-reviewer-patterns, asm-forge, review-pipeline, perf-pipeline, cache-correctness-review, security-reviewer, etc.).
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.