The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources that can contain natural language instructions, potentially influencing agent behavior.
Ingestion points: The skill retrieves raw user prompts and conversation history using commands such as cawplan session conversation, cawplan session human-inputs, and cawplan session user-human-inputs as described in SKILL.md.
Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between the agent's instructions and the fetched data.
Capability inventory: The skill has access to the Bash tool, allowing it to execute CLI commands and potentially interact with the file system or project metadata.
Sanitization: There is no evidence of sanitization or validation of the fetched conversation text before it is returned to the agent's context.
[COMMAND_EXECUTION]: The skill frequently uses the Bash tool to execute cawplan CLI commands with interpolated parameters.
The workflow involves constructing shell commands using variables for dates, user IDs, and product IDs (e.g., cawplan session product-sessions --product-id <pid> --date <date>). While these are functional requirements for a management tool, they represent the primary execution vector for the skill.

cawplan-coding-insights