gcc

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local shell script, scripts/gcc_init.sh, to set up the necessary directory structure and template files. The script is restricted to basic file system operations (mkdir, cat) and does not perform network requests or require elevated privileges.
  • [INDIRECT_PROMPT_INJECTION]: As a context management tool, the skill reads historical summaries and logs from the .GCC/ directory to recover state. This represents a potential surface for indirect prompt injection if malicious data is committed to the project history, but this is a low-risk, inherent characteristic of memory-persistence skills.
  • [DATA_EXPOSURE_EXFILTRATION]: File operations are strictly scoped to the project root and the specific .GCC/ directory. No sensitive paths (such as SSH keys or environment files) are accessed, and there are no network capabilities defined in the skill to exfiltrate data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 02:37 PM
Security Audit — agent-trust-hub — gcc