The Agent Skills Directory

[COMMAND_EXECUTION]: The skill orchestrates multiple specialized tools and agents, such as codebase-locator, codebase-analyzer, and context7, to perform research. It also invokes internal sub-skills like desplega:reviewing and desplega:learning for workflow management.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it reads external files from the codebase and research documents to generate implementation plans.
Ingestion points: The skill reads all mentioned files and directories identified during research tasks, as well as the local ~/.agentic-learnings.json file.
Boundary markers: There are no explicit instructions to the agent to treat external file content as untrusted or to use specific delimiters.
Capability inventory: The skill has the ability to write plan files to the thoughts/ directory and propose commands for implementation.
Sanitization: The skill does not describe any sanitization or validation logic for the content read from files.

planning