launch: submit to awesome-claude-code (Security) and awesome-ai-security #209

Part of the 2026-10-24 launch plan (step 5, directories). Open these PRs the day after the Show HN, not before — the lists want projects with some signal.

Targets (checked 2026-10-24)

1. hesreallyhim/awesome-claude-code (56k stars, pushed today) — ## Security section (README line ~461). Fit: direct — the tool scans CLAUDE.md/skills and ships a Claude Code PostToolUse hook in integrations/claude-code/. Follow that repo's CONTRIBUTING (it uses a submission form/issue in some periods — check first).
2. ottosulin/awesome-ai-security (1.6k stars, pushed today) — ### MCP Security (line ~270) or ### Safety Classifiers & Prompt Injection Detection (line ~355). Fit: direct. Format is * [name](url) - _one sentence._.
3. punkpeye/awesome-mcp-servers ### 🔒 Security (line ~4121, 97k stars): evaluated and not recommended — that section lists MCP servers; injection-scanner is a CLI (mcp-scan is not listed there either). Revisit only if an MCP server mode ships.

Also worth a line: FonduAI/awesome-prompt-injection (0.7k, active) and corca-ai/awesome-llm-security (1.7k, last push 2025-08 — likely unmaintained, skip).

One-sentence description to reuse

injection-scanner — Rust static scanner for prompt-injection patterns in agent-facing files (CLAUDE.md, skills, MCP tool descriptions, RAG docs); SARIF output, pre-commit hook, published recall numbers including the misses.

Done when

• 2 PRs opened and linked in the tracking issue; merged or not is outside our control.

Effort: ~1 h.