plan-perf-audit
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data sources including application source code, database EXPLAIN logs, and performance metrics from monitoring services. This creates a surface for indirect prompt injection if these resources contain instructions designed to manipulate the agent's performance analysis or planning logic.\n
- Ingestion points: Application source code, database performance logs (EXPLAIN), and performance monitoring reports (Sentry/Lighthouse).\n
- Boundary markers: No specific delimiters or warnings for embedded instructions are defined for the data processed during the audit.\n
- Capability inventory: The skill is intended for planning, but it operates in an environment where the agent may have access to shell and file system tools.\n
- Sanitization: The instructions do not specify any validation or sanitization procedures for the data ingested from external sources.\n- [COMMAND_EXECUTION]: The skill requires the agent to run development commands such as
npm run build,npm audit, andlighthouseto establish performance baselines. These are standard diagnostic commands used within the intended scope.\n- [EXTERNAL_DOWNLOADS]: The skill usesnpx lighthouse, which facilitates the download and execution of the Lighthouse auditing tool from a well-known service. This is documented as a standard measurement procedure.
Audit Metadata