The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts as part of its workflow. Specifically, it calls risk_quantifier.py and compliance_tracker.py located in a relative directory structure (../../../skills/ciso-advisor/scripts/). This is a functional requirement of the skill's logic for quantifying risk and tracking compliance.
[PROMPT_INJECTION]: The skill defines a command /cs:ciso-review <plan> that ingests untrusted user data via the <plan> argument.
Ingestion points: Untrusted data enters the agent context through the <plan> parameter in SKILL.md.
Boundary markers: There are no explicit delimiters or system instructions provided to the agent to treat the content of the plan as data rather than instructions, which could allow a malicious plan to influence the agent's security verdict.
Capability inventory: The skill has the capability to execute shell commands (Python scripts) as defined in the Workflow section of SKILL.md.
Sanitization: No sanitization or validation logic is visible for the ingested plan data before it is processed by the agent or used in the review workflow.

ciso-review