git-commit

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs local shell operations using git and the GitHub CLI (gh) to stage changes, create commits, sign tags, and generate releases. It also executes standard developer tools like npm, bun, and cargo as part of the verification process.
  • [INDIRECT_PROMPT_INJECTION]: The skill reads project manifests and task runner configurations to identify and execute verification commands, which represents a potential attack surface if the agent is used on a maliciously crafted repository.
  • Ingestion points: Project manifest files (e.g., package.json, Makefile, justfile) and the output of git diff command.
  • Boundary markers: None specified for input data.
  • Capability inventory: Execution of arbitrary shell commands defined in manifest files, file system modifications (git staging), and network access via the GitHub CLI.
  • Sanitization: No validation or filtering is performed on the commands retrieved from the project files before they are executed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 10:05 AM