The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local configuration files (~/.agentic-learnings.json) and design documentation (thoughts/*/design-docs/) to inform its implementation plans. * Ingestion points: Local learning files and project-specific design documents (SKILL.md). * Boundary markers: None explicitly defined within the skill instructions. * Capability inventory: The skill can write implementation plans, specify and execute shell commands for verification, and delegate tasks to sub-agents (SKILL.md). * Sanitization: No specific sanitization or filtering of the ingested context is mentioned. * Context: This risk is mitigated by the skill's design, which requires human verification at each phase and mandates that the final implementation occur in a separate session (Rule 10).
[COMMAND_EXECUTION]: The skill instructs the agent to include and execute automated verification commands such as 'make test' or 'npm test' as part of the implementation success criteria. This is a standard functional requirement for automated development planning and is supported by human-in-the-loop confirmation requirements throughout the process.

planning