agent-readiness
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of agent session logs from local directories (e.g.,
~/.claude/projects/). - Ingestion points:
scripts/analyze_sessions.pyiterates through and parses JSONL files containing transcripts of agent interactions. - Boundary markers: No specific delimiters are used to wrap or isolate log content during parsing.
- Capability inventory: Across all scripts, capabilities are restricted to local file system read operations and writing a report file to
/tmp/agent-readiness-report.html. No network or subprocess capabilities are present. - Sanitization:
scripts/render_html_report.pyutilizeshtml.escape()to sanitize all content before it is rendered in the final report, mitigating potential injection risks from processed log data. - [EXTERNAL_DOWNLOADS]: Informational links to the developer's GitHub repositories (
supatest-ai/supa-skillsandsupatest-ai/aiden) are present inSKILL.md. These links are for user reference and are not accessed or executed by the skill's scripts.
Audit Metadata