tracer-worktrees

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions utilize shell commands for repository state inspection and workspace manipulation. It specifically uses git utilities for branch detection, directory validation, and worktree management.
  • [EXTERNAL_DOWNLOADS]: Step 3 of the process involves setting up the project environment by installing dependencies through standard package managers such as npm, pip, cargo, and go. These tools download external software from established public registries to initialize the feature workspace.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from the local repository environment, including branch names and directory structures. While this ingestion represents a potential attack surface for malformed metadata to influence command parameters, the operations are limited to standard development lifecycle tasks.
  • Ingestion points: Branch names retrieved via git symbolic-ref and path information from git rev-parse (SKILL.md).
  • Boundary markers: None explicitly defined for shell command interpolation.
  • Capability inventory: File system modification, directory creation, and network-enabled package installation (SKILL.md).
  • Sanitization: Standard shell execution is used without explicit input escaping for branch names.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 03:34 PM
Security Audit — agent-trust-hub — tracer-worktrees