skills/basher83/lunar-claude/git-sync/Gen Agent Trust Hub

git-sync

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes standard git commands (git fetch, git pull, git status) to manage repository state. These operations are restricted to the local workspace and its configured remotes.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a minimal attack surface for indirect injection by reading repository metadata.
  • Ingestion points: Reads git remote get-url, git branch --show-current, and git status --porcelain in SKILL.md.
  • Boundary markers: None explicitly used in prompt interpolation; however, the data is used for control flow rather than instruction generation.
  • Capability inventory: File system modifications via git pull and git merge in SKILL.md.
  • Sanitization: The skill implements strict state checks, such as aborting if the working tree is dirty (git status --porcelain having any output), which prevents common injection-through-filename attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 04:53 PM
Security Audit — agent-trust-hub — git-sync