ast-grep

Pass

Audited by Gen Agent Trust Hub on Oct 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill includes installation scripts (install.sh and install.ps1) that download the ast-grep binary from its official repository on GitHub (github.com/ast-grep/ast-grep). These downloads target a well-known service and organization.
  • [COMMAND_EXECUTION]: The Python helper script (scripts/ast_grep_helper.py) executes the ast-grep binary using subprocess.run. The command is passed as a list, and the script includes logic to resolve the binary path across different environments (cached local binary, PATH, OMO runtime, or Homebrew).
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external code files, which constitutes a potential attack surface for indirect prompt injection. However, the skill implements a mandatory evidence chain for safety: it identifies ingestion points via the paths argument, performs offline validation of patterns to catch common errors or injections, and limits its capabilities to AST-aware code transformations.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 17, 2026, 06:56 AM