The Agent Skills Directory

[SAFE]: The skill implements a standard planning workflow. It uses interactive prompts to gather requirements and generates documentation in the docs/plans/ directory. No evidence of obfuscation, credential harvesting, or unauthorized data exfiltration was found.
[COMMAND_EXECUTION]: The skill refers to executing project-specific test commands during the implementation phase. These commands are triggered by the agent based on the project's own configuration (e.g., npm test, pytest) as part of the intended development workflow.
[PROMPT_INJECTION]: The skill uses instructional markers like 'CRITICAL' and 'IMPORTANT' to enforce plan consistency and testing requirements. These are benign behavioral constraints designed to ensure the quality of the generated plan rather than attempts to bypass AI safety guidelines.
[DATA_EXPOSURE]: The skill reads local project metadata and file structures via the spring-explore skill to generate relevant plans. This data remains within the local agent context and is used solely for the stated purpose of implementation planning.

spring-planning