[Target] Add Puppeteer MCP to the Safety Index

[KryptosAI]

Task

Run MCP Observatory against Puppeteer MCP (@anthropic/mcp-server-puppeteer) and add a Safety Index target.

Why this matters

Puppeteer MCP enables agents to control headless Chrome — navigate pages, click elements, take screenshots. The security surface includes arbitrary browser automation, potential SSRF via page navigation, and access to rendered DOM content.

Steps

1. Run the scan: npx @kryptosai/mcp-observatory test npx -y @anthropic/mcp-server-puppeteer
2. Generate the report: npx @kryptosai/mcp-observatory report --run <artifact.json> --format markdown
3. Add artifacts to docs/safety-index/artifacts/puppeteer-mcp-server.{json,md}
4. Add entry to docs/safety-index/targets.json
5. Update the safety index table in docs/mcp-server-safety-index.md
6. Run npm run safety-index and npm test

Safe mode

The tool listing works without a running browser instance. The scan runs in metadata-only safe mode by default.

Reference

See PR #209 (UI5 MCP) for the expected output format. 15-20 minutes.