meeting-insights-analyzer
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists exclusively of instructions (SKILL.md) and metadata (.skill-meta.json). It does not include any Python scripts, JavaScript files, shell scripts, or binaries.
- [DATA_EXPOSURE]: The skill is designed to analyze meeting transcripts stored in a user-specified local directory. It does not attempt to access sensitive system directories, configuration files, or credentials. There are no hardcoded secrets or API keys.
- [PROMPT_INJECTION]: The instructions do not contain patterns intended to bypass AI safety filters or override system prompts. The language is focused on communication analysis and feedback.
- [DATA_EXFILTRATION]: No network operations (such as curl, wget, or fetch) are present. The skill does not attempt to send any analyzed data to external servers.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted meeting transcript data. However, the risk is negligible because the skill lacks the capabilities (e.g., shell access, network requests, file writing) required to execute a malicious payload. Ingestion points: Reads transcript files (.txt, .md, .vtt, .srt, .docx) from local folders. Boundary markers: None defined for transcript content. Capability inventory: None. The skill has no capability to execute code or perform network requests. Sanitization: None specified.
Audit Metadata