The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using cargo audit and cargo geiger. These commands are standard for auditing Rust projects and do not include suspicious parameters or redirection to sensitive locations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external security tools which process project metadata, creating a potential surface for indirect prompt injection if the project being audited contains malicious content in crate names or vulnerability descriptions.
Ingestion points: The skill reads cargo-audit-results.json and cargo-geiger-results.json to generate summaries.
Boundary markers: No specific boundary markers or instructions to treat tool output as untrusted are present in SKILL.md.
Capability inventory: The agent has the capability to execute shell commands and write to the local file system as part of the auditing process.
Sanitization: There is no evidence of sanitization or filtering of the tool output before the agent processes and summarizes it.

sast-cargo-audit