The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves analyzing untrusted external data, specifically code diffs and pull request content. This creates a surface for indirect prompt injection where an attacker could potentially embed malicious instructions within code comments or commit messages to influence the agent's security findings.
Ingestion points: The skill fetches untrusted data using gh pr diff and git diff as described in Step 2 of the instructions.
Boundary markers: The instructions do not define specific delimiters or instructions for the agent to ignore potentially malicious text embedded within the code diffs.
Capability inventory: The skill possesses the ability to execute shell commands (via git and gh) and write files to the local file system (security-findings.json).
Sanitization: There is no evidence of sanitization or filtering performed on the input diff content before it is processed by the reasoning engine.

commit-security-scan