jmb-implement-frontier

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted data from GitHub issue bodies to determine orchestration logic and to construct prompts for subagents.
  • Ingestion points: The skill fetches data from GitHub repositories using gh api and specifically parses the issue body for a Spec: line and various criteria as described in SKILL.md (Step 2).
  • Boundary markers: No explicit delimiters or instructions are provided to separate the untrusted data from the agent's system instructions.
  • Capability inventory: The skill possesses the capability to execute shell commands (git worktree, gh api), modify repository state, and spawn background subagents.
  • Sanitization: The instructions do not define any sanitization or validation steps for the content retrieved from issue bodies before it is interpolated into logic or subagent prompts.
  • [COMMAND_EXECUTION]: The skill makes extensive use of local shell commands to manage the git environment and interface with GitHub.
  • Evidence: SKILL.md contains instructions to execute git branch, git status, git pull, git worktree add, and various gh CLI commands to manage issues and pull requests.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 07:02 PM