git-workflow-skill
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to write execution metrics and outcomes to a local file located at
~/.claude/skill-analytics/last-outcome-git-workflow-skill.jsonto track usage statistics like commits created and PRs opened. - [INDIRECT_PROMPT_INJECTION]: The skill provides templates for shell commands (e.g.,
gh pr create --title "..." --body "...") that interpolate user-supplied strings. This represents a potential injection surface if the input contains shell metacharacters and is not properly sanitized by the agent before execution. - Ingestion points: User-provided descriptions for commits, PR summaries, and ticket IDs.
- Boundary markers: None explicitly defined in the shell command examples.
- Capability inventory: Shell command execution via
git,gh(GitHub CLI), andnpm/npxacross various workflow steps. - Sanitization: No specific sanitization instructions are provided for handling special characters in user input.
Audit Metadata