skills/koolamusic/claudefiles/kickoff/Gen Agent Trust Hub

kickoff

Pass

Audited by Gen Agent Trust Hub on Jun 25, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill performs a 'Codebase scan' by reading local files such as package.json, Cargo.toml, go.mod, and git history. This is used solely to detect the project's tech stack and existing state for initialization purposes. No data is transmitted externally.
  • [COMMAND_EXECUTION]: The skill uses several local shell commands to manage the project structure:
  • ls for codebase exploration.
  • mkdir -p .project/ for directory creation.
  • git add and git commit to version control the generated documentation. These are standard operations for the skill's stated purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local files (e.g., README.md, package.json) and user input to generate PROJECT.md and ROADMAP.md. While this is an ingestion surface, the skill uses the information for template-based document generation and requires user iteration/approval before final commitment, mitigating the risk of automated obedience to embedded instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 25, 2026, 11:26 AM
Security Audit — agent-trust-hub — kickoff