The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes data from external documentation files and user input to drive its decision logic and file operations.
Ingestion points: The skill reads documentation files (path provided via doc_path) and accepts direct user input for skill names, categories, and custom actions.
Boundary markers: None. There are no delimiters or instructions provided to the agent to disregard instructions embedded within the processed documentation or user inputs.
Capability inventory: The skill has access to Write, Edit, and Bash tools, allowing it to modify the file system and execute shell commands.
Sanitization: None. The skill does not specify any validation, escaping, or filtering of the ingested content before it is written to new files or used in command contexts.
[COMMAND_EXECUTION]: The skill uses the Bash tool and includes an "Other" option (Option 8) that prompts for and executes custom actions. This capability, combined with the lack of input sanitization, creates a risk where malicious instructions in the processed documentation could influence the execution of shell commands.

fix-decision-router