Idea Flow Orchestrator
Chooses the right next skill and publishing action for a raw prompt, existing idea, research request, or pro-candidate request.
- Classify the user input.
- Ask only the missing questions.
- End with a tool plan or blocker.
These are the FreeIdeaStore-native playbooks agents should follow before creating, critiquing, researching, validating, prototyping, refining, or promoting an idea. They are also exposed through MCP as list_idea_skills, get_idea_skill, and apply_idea_skill. Local Codex or Claude skills can follow these playbooks, but the canonical published versions live here and in FIS MCP.
Chooses the right next skill and publishing action for a raw prompt, existing idea, research request, or pro-candidate request.
Turns a vague idea into enough structured context to publish a useful first page.
Turns the first prompt or short interview answers into a coherent generic FreeIdeaStore public idea document.
Finds weak assumptions, safety issues, regulation, cost traps, trust gaps, and reasons to trash or narrow the idea.
Finds direct competitors, indirect substitutes, and manual workarounds before an idea is treated as novel.
Turns public evidence, competitor notes, user signals, and source-backed observations into useful idea history.
Merges useful history into a clearer canonical public idea document while keeping risks and caveats visible.
Finds sharper customer, market, product, or format alternatives when the original version is too broad or risky.
Turns an idea into a cheap validation plan that can prove, narrow, pivot, or trash it quickly.
Defines the smallest prototype that can test the idea's core workflow without pretending the full product exists.
Decides whether an idea deserves ProIdeaStore diligence, needs more refinement, should pivot, or should be parked.