The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to perform development operations including software builds (npm run build), linting, and unit testing (npm test). It also handles project deployment tasks for platforms like Vercel and Railway based on the generated architecture.
[EXTERNAL_DOWNLOADS]: The web_search and fetch_page tools are used during the 'Discovery' phase to research competitors and technology stacks. This external information directly influences the project's technical decisions.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks as it ingests and processes untrusted data from the web and user-provided project descriptions to generate implementation stories and code.
Ingestion points: External research data gathered by the 'Analyst' role and initial project ideas received by the 'Commander' role.
Boundary markers: The prompt instructions do not explicitly implement delimiters or specific isolation instructions to prevent instructions embedded within external data from being executed by the agents.
Capability inventory: The system possesses high-privilege tools including arbitrary shell execution (Bash), file modification (Write, Edit), and deployment capabilities.
Sanitization: No explicit sanitization or filtering logic was identified to process external inputs before they are interpolated into the context of the Builder or Architect agents.

aiox-autopilot