workflow-code-to-spec

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project-level files such as source code, workflows.json, and project-config.json. This creates a surface where instructions embedded in these external sources could theoretically influence the agent's actions.
  • Ingestion points: Accesses project source code, .claude/workflows.json, and docs/project-config.json (File: SKILL.md).
  • Boundary markers: The workflow relies on the agent's context window and model reasoning rather than explicit data delimiters.
  • Capability inventory: Includes file system writing (tmp/reports/, goal.md), spawning of sub-agents (spawn_agent), and local script execution.
  • Sanitization: Incorporates mandatory quality gates and review protocols such as $work-item --mode=review and $why-review to validate findings.
  • [COMMAND_EXECUTION]: The skill facilitates the execution of a project-specific script using node .claude/scripts/lib/workflow-skill-contract.cjs --write. This operation is used for synchronizing the workflow contract with the project's registry.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 10:12 AM