What problem does it solve?
Long planning conversations converge on a clear next task, but handing that work to the /goal autonomous loop requires manually re-summarizing every decision, path, and constraint. This Skill distills the current conversation into one ready-to-run /goal prompt so nothing load-bearing is lost in the handoff.
Core Features & Use Cases
- Conversation-to-goal synthesis: Mines the discussion for decisions, exact file paths, commands, constraints, and current state, then composes a directive a cold agent can execute without shared memory.
- Measurable completion conditions: Every generated prompt ends with an evaluator-checkable end state, since the
/goal loop stops only when a small evaluator model confirms the condition from the transcript.
- Enforced length validation: Ships a Python script that verifies the prompt argument stays under the 4000-character
/goal limit before output, reporting headroom or overage.
- Use Case: After a long debugging session where you agreed to add a default-off feature flag with pin-neutral tests, ask for a goal prompt and receive a single fenced block containing the directive, context, constraints, and completion criteria ready to paste into
/goal.
Quick Start
Ask Claude to turn this conversation into a /goal prompt for the work we just agreed on.
