audit-skill-lifecycle
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content by scanning all skills, agents, commands, and reference files within a plugin directory to perform its multi-dimensional audit.
- Ingestion points: The 'Discovery' step (Step 1) scans the plugin structure to build an inventory of external markdown, JSON, YAML, and data files for analysis.
- Boundary markers: The instructions lack explicit boundary markers or directives to treat the content of audited files as untrusted, increasing the risk of the model following instructions embedded within the data it is supposed to be auditing.
- Capability inventory: The skill has the capability to write multiple report files to the
.plugin-creator/audits/directory and can delegate complex semantic tasks to Tier 3 'Specialist Delegation' agents. - Sanitization: No sanitization, filtering, or escaping of the ingested content is described before it is processed by the auditor or its delegated specialist agents.
Audit Metadata