health
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a multi-stage Bash script to collect project metrics, file counts, and configuration details across the local environment and the home directory.
- [DATA_EXFILTRATION]: The skill accesses sensitive local project data, including settings.local.json (which may contain configuration secrets), historical conversation logs (.jsonl), and architectural memory files (MEMORY.md) to perform its audit.
- [DATA_EXFILTRATION]: During the MCP Live Check, the skill attempts to verify the presence of API keys by echoing the first 5 characters of specific environment variables.
- [PROMPT_INJECTION]: The skill ingests and processes untrusted data from external SKILL.md files and user conversation history, which creates a surface for indirect prompt injection. It includes mitigation instructions to sub-agents to treat this content as untrusted data and uses redaction for credentials before delegating tasks.
Audit Metadata