find-replace

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is a legitimate collection of documentation and command templates for the sd utility, designed to help developers perform string replacements more easily than with traditional tools like sed.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines an interface for the agent to modify project files based on search patterns.
  • Ingestion points: The skill processes content from local project files, including source code files (.ts, .js, .tsx), configuration files (.env, .yaml, .json), and text documents.
  • Boundary markers: While the skill does not provide platform-level prompt boundaries, it explicitly teaches the use of regex word boundaries (\b) and quoting to ensure replacements are scoped correctly and do not affect unintended code segments.
  • Capability inventory: The skill uses sd for in-place file writing, and utilizes ripgrep (rg) and fd for discovery and content inspection.
  • Sanitization: The documentation strongly advocates for manual verification steps, recommending that users/agents preview changes with rg -r and review output via git diff before committing modifications.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 01:18 PM