The Agent Skills Directory

[SAFE]: The skill provides legitimate functionality for analyzing Rust codebases using standard LSP operations such as prepareCallHierarchy, incomingCalls, and outgoingCalls. No malicious patterns were detected.
[DATA_EXPOSURE]: The skill uses LSP, Read, and Glob tools. While these allow access to the local project files, they are within the expected scope for a code analysis tool. No network tools or external data exfiltration patterns were identified.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it processes untrusted data (source code) through the Read and LSP tools. However, since the focus is on structural call hierarchy rather than interpreting natural language comments, the risk is minimal.
Ingestion points: Files read via Read and symbols processed via LSP (SKILL.md).
Boundary markers: None explicitly defined in instructions.
Capability inventory: LSP, Read, Glob (SKILL.md).
Sanitization: None specified for file content.

rust-call-graph