ripgrep

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the execution of the rg (ripgrep) binary for filesystem search operations. It includes examples of piping output to other utilities such as jq for JSON processing and xargs sed for find-and-replace workflows.
  • [EXTERNAL_DOWNLOADS]: Mentions installation via trusted system package managers (apt, dnf) and provides links to the official ripgrep GitHub repository (BurntSushi/ripgrep) and documentation.
  • [PROMPT_INJECTION]: The skill describes a tool that processes external, untrusted file contents, which inherently creates a surface for indirect prompt injection. 1. Ingestion points: File contents read by rg during recursive searches as described in SKILL.md and cheatsheet.md. 2. Boundary markers: None identified in the provided command examples. 3. Capability inventory: Filesystem read access (rg), potential filesystem write access (sed -i via xargs), and terminal output. 4. Sanitization: No specific sanitization or validation of search results is defined within the skill documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 01:37 PM