The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external repositories which serves as a significant attack surface for indirect prompt injection.
Ingestion points: The skill reads source files, configurations, package manifests, and CI/CD pipelines across the entire codebase (SKILL.md).
Boundary markers: There are no explicit instructions or delimiters defined to separate untrusted repository content from the agent's instructions during the analysis phase.
Capability inventory: The agent is granted powerful tools including Bash, Write, and Edit, which could be abused if the agent follows instructions found within a repository being analyzed.
Sanitization: No sanitization or validation of the input data is specified before it is processed by the LLM.

project-profiler