The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted content from a local codebase, creating a surface for indirect prompt injection where malicious instructions could be hidden in the code being audited.
Ingestion points: The skill reads arbitrary files using the Read, Grep, and Glob tools to perform analysis.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to distinguish between its operational instructions and the potentially untrusted content of the audited files.
Capability inventory: The skill has access to Bash for running linters and coverage tools, which could be abused if an attacker succeeds in an indirect prompt injection.
Sanitization: No sanitization or filtering of the codebase content is specified before it is processed by the agent.

code-auditor