project-init
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell commands and Node.js maintenance scripts within the project framework. Evidence: Calls to scripts like .claude/skills/project-config/scripts/project-config-help.cjs and .claude/hooks/tests/run-all-tests.cjs.
- [DYNAMIC_EXECUTION]: Utilizes node -e to perform runtime evaluation of JavaScript snippets for environment assessment. Evidence: Usage of node -e to load hooks from .claude/hooks/lib/session-init-helpers.cjs and agent-files-state.cjs in Phase 0 and Phase 4.
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from the repository to derive configuration, creating a potential vector for instructions hidden in data. * Ingestion points: Reads repository metadata, manifests, and existing user-authored CLAUDE.md files. * Boundary markers: No specific boundary delimiters are defined for the ingestion of project metadata or files in the provided instructions. * Capability inventory: Capable of writing project configuration, updating root context files, executing shell commands, and spawning sub-agents. * Sanitization: Validates the resulting project configuration against the project-config-schema.cjs validator.
Audit Metadata