What problem does it solve? Users often have vague, messy, or under-specified ideas that fail as prompts for AI agents, producing weak or misaligned results. This Skill infers the strongest useful task inside a rough seed and expands it into a complete, execution-ready prompt with explicit assumptions, constraints, deliverables, and checkable completion criteria. ## Core Features & Use Cases - Seed-to-prompt forging: Infers the real goal from a vague or misspelled seed and labels every inferred assumption explicitly instead of baking in silent guesses. - Dual templates with a shared contract: Uses a repo-agent template (with a mandatory workspace hygiene block) for coding targets and a non-code template with evidence/citation discipline for research and writing tasks. - Quality gate self-check: Validates every forged prompt against a rubric covering concrete goals, bounded scope, specific deliverables, and verifiable "done when" criteria before returning it. - Use Case: A user pastes "fix the flaky tests in my repo" and receives a full Claude Code prompt including repo context sections, a read-only-first hygiene block, forbidden-by-default destructive operations, and test-pass completion criteria. ## Quick Start Ask the agent to forge your rough idea into a master prompt, for example: "Forge this into a master prompt for Claude Code: add rate limiting to the API."
