watzup
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from git diffs, logs, and prompt ledgers.
- Ingestion points: The skill reads uncommitted changes via
git diff, commit history viagit log, and session history fromledger.md(SKILL.md). - Boundary markers: The skill instructions mandate the use of HTML-escaping for all placeholders and safe link protocols in the final report (SKILL.md).
- Capability inventory: The agent has access to shell commands (
git,ls,node) and file system write operations (SKILL.md). - Sanitization: Explicit instructions for HTML entity encoding and URI scheme filtering are provided. The generated report includes a strict Content Security Policy (CSP) meta tag (references/session-report-template.html).
- [COMMAND_EXECUTION]: The skill uses shell commands for repository inspection and reporting.
- Evidence: Execution of
git status,git diff,git log,ls, andnode .claude/scripts/open-report.cjs(SKILL.md).
Audit Metadata