The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a vulnerability to indirect prompt injection due to its core functionality of processing external data.
Ingestion points: Technical tickets and research documents located in the thoughts/research/ directory.
Boundary markers: The skill lacks explicit instructions or markers to distinguish between system-level planning instructions and untrusted content within source materials.
Capability inventory: The skill utilizes the Task tool to spawn sub-agents and the Write/Edit tools to modify the codebase based on its planning outputs.
Sanitization: No mechanisms for sanitizing or validating input from tickets or research documents are defined, potentially allowing embedded instructions to influence the agent's behavior.

implementation-planning