context-analyzer
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's primary functionality is limited to reading and analyzing local project files (e.g., Nuxt directories, Better-Auth configurations, and package.json) for the purpose of symbol resolution and logic tracing. It does not include commands for network exfiltration, persistence, or privilege escalation.
- [PROMPT_INJECTION]: The skill ingests untrusted data by reading project files with extensions .ts, .vue, .json, and .md (Ingestion points: SKILL.md). It lacks explicit boundary markers or sanitization instructions to prevent the agent from obeying instructions embedded within the analyzed code. Capability inventory: The skill utilizes context-gathering capabilities to resolve symbols and trace logic flows. Sanitization: No sanitization or validation of the ingested code content is specified.
Audit Metadata