sh-style-guide
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute shell-based linting and testing tools including
shellcheck,shfmt, andshunit2. These are primarily used to verify code compliance with POSIX standards for files identified as POSIX shell scripts.\n- [DYNAMIC_EXECUTION]: The skill uses Python to execute internal repository scripts, specificallyscripts/resolve_style_guides.pyandscripts/validate_trigger_matrix_sync.py. These scripts are invoked to dynamically determine which style guide rules to apply based on the specific files modified in a pull request, creating an execution dependency on local repository content.\n- [INDIRECT_PROMPT_INJECTION]: As a code review and refactoring skill, it processes untrusted source files, which presents a surface for indirect prompt injection. This is evaluated as a vulnerability surface common to static analysis agents.\n - Ingestion points: The agent reviews the raw contents of files matching
.shextensions, shell shebangs, or CI workflows containing shell blocks.\n - Boundary markers: No specific delimiters or 'ignore-instructions' warnings are defined to isolate untrusted script content from the agent's internal reasoning context during the review process.\n
- Capability inventory: The agent is empowered to run shell commands, linters, and unit tests (
shunit2) against the code it ingests.\n - Sanitization: The skill does not implement any sanitization, filtering, or escaping for the data extracted from reviewed files before they are processed.
Audit Metadata