constitution

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns such as obfuscation, unauthorized data exfiltration, or remote code execution were found. The skill's behavior matches its stated purpose of project documentation and principle management.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes untrusted data from project files to draft a constitution.
  • Ingestion points: SKILL.md (Workflow Step 1) reads project context from files including README.md, docs/, and configuration files like package.json.
  • Boundary markers: None identified in the workflow instructions.
  • Capability inventory: The skill can read local files and write/edit the CLAUDE.md file. It does not have network access or arbitrary shell execution capabilities.
  • Sanitization: No specific content validation or escaping is applied to the data ingested from external project files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 02:59 AM
Security Audit — agent-trust-hub — constitution