The Agent Skills Directory

[COMMAND_EXECUTION] (HIGH): The skill defines the run_command tool in its metadata. Although no malicious scripts are currently present, providing execution capabilities to an agent that processes untrusted external data is a high-risk configuration.
[PROMPT_INJECTION] (HIGH): The skill is highly vulnerable to Indirect Prompt Injection (Category 8).
Ingestion points: Data enters the context through user interviews (via socratic_interrogation.md) and file analysis (via read_file).
Boundary markers: There are no instructions for the agent to use delimiters or ignore instructions embedded within the data it processes.
Capability inventory: The agent has access to run_command, write_file, and edit_file, allowing for persistent system changes or execution of arbitrary code if triggered by an injection.
Sanitization: No input validation or sanitization logic is described to handle potentially malicious payloads in PRDs or interview transcripts.
[PROMPT_INJECTION] (LOW): The skill's identity instructions in SKILL.md ("You do not take orders; you define outcomes") attempt to harden the agent's persona against direct manipulation, but this does not mitigate the indirect injection risks mentioned above.

resonance-product