skills/opengovsg/formsg-ai/split-prs/Gen Agent Trust Hub

split-prs

Pass

Audited by Gen Agent Trust Hub on Aug 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the installation of the gh-stack extension from the official GitHub organization repository. This is a trusted source for developer tooling and recognized as a well-known service.- [COMMAND_EXECUTION]: The skill utilizes the local shell to execute git and gh commands for managing repository state, commits, and pull requests. These actions are directly relevant to the skill's purpose and are protected by a required user approval gate.- [PROMPT_INJECTION]: The skill processes data from git diffs, logs, and pull request bodies, which creates an indirect prompt injection surface. The workflow mitigates this by requiring the user to review and approve the agent's plan before any destructive or network-facing actions occur.
  • Ingestion points: Git diffs, log history, and existing pull request descriptions (SKILL.md).
  • Boundary markers: The instructions do not define specific delimiters to isolate ingested data from the agent's instructions.
  • Capability inventory: The agent uses shell-level execution for repository management tools (git, gh).
  • Sanitization: No explicit sanitization or filtering is applied to the content retrieved from the git repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 9, 2026, 01:24 PM