scripted-formatting-delegation
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute various CLI-based formatting tools including
ruff,gofmt,cargo fmt, andclang-formatto modify local source files. \n- [EXTERNAL_DOWNLOADS]: The instructions recommend usingnpx -yto runprettierand@biomejs/biome, which involves fetching these packages from the NPM registry at runtime. These are well-known and widely used development services. \n- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to run formatting tools on repository content. While these tools are used for their intended purpose, the agent is acting upon potentially untrusted data in the codebase. \n - Ingestion points: Local source files (Python, JS, TS, JSON, etc.) and repository configuration files. \n
- Boundary markers: None present in the instructions. \n
- Capability inventory: Shell command execution via
run_commandto invoke formatters. \n - Sanitization: No explicit sanitization or validation of the files being formatted is mentioned.
Audit Metadata