spec-implement

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill extracts and executes shell commands for environment setup, testing, and linting that are dynamically parsed from the issue-to-pr-workflow.md file within the repository.
  • [PROMPT_INJECTION]: The skill adopts instructions from project-level files such as CLAUDE.md, AGENTS.md, and coding-rules.md as mandatory enforcement rules, which may influence agent behavior.
  • [PROMPT_INJECTION]: The skill exhibits an architectural surface for indirect prompt injection. * Ingestion points: Content is read from GitHub issue bodies and local specification files within the .specs/ directory. * Boundary markers: No explicit delimiters or instructions to disregard nested prompts are implemented when processing these external data sources. * Capability inventory: The agent has the authority to modify the local file system, perform git operations (commit/push), and execute arbitrary shell commands defined in the project's workflow. * Sanitization: Data from GitHub issues and local specification files is used directly to guide the implementation loop without pre-processing or validation.
  • [EXTERNAL_DOWNLOADS]: The skill uses the official GitHub CLI (gh) to communicate with GitHub services for viewing issue context and creating pull requests.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 07:26 AM