- Home
- Skills
- Orchestra Research
- Ai Research Skills
- Brainstorming Research Ideas
brainstorming-research-ideas_skill
- TeX
5.2k
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill orchestra-research/ai-research-skills --skill brainstorming-research-ideas- SKILL.md18.7 KB
Overview
This skill guides researchers through ten structured ideation frameworks to move from vague curiosity to concrete, defensible research directions. It packages practical workflows for problem framing, analogy transfer, failure analysis, and composition so you can generate and rank high‑impact ideas quickly. Use it to explore new areas, pivot projects, or validate half‑formed concepts.
How this skill works
The skill inspects your starting point (a one‑sentence focus, capability, or failure) and applies complementary lenses: problem‑first vs solution‑first, abstraction ladder, tension hunting, cross‑pollination, change detection, boundary probing, simplicity testing, stakeholder rotation, composition/decomposition, and the two‑sentence clarity test. For each lens it provides short workflows and self‑checks that turn intuition into testable hypotheses and candidate proposals.
When to use it
- Starting a new research direction and needing structured exploration
- Stuck on a project and wanting fresh angles or higher leverage moves
- Evaluating whether a half‑formed idea has real impact potential
- Preparing a focused brainstorming session with collaborators
- Transitioning between fields and seeking underexplored entry points
Best practices
- Always state your idea in one sentence before applying a lens
- Combine multiple frameworks (e.g., tension hunting + analogy) for richer candidates
- Run the two‑sentence test early to filter unclear or low‑impact directions
- Document assumptions explicitly and test which have actually changed
- Prioritize ideas that name specific stakeholders and measurable outcomes
Example use cases
- Turn a novel model capability into two concrete, stakeholder‑driven applications
- Revisit an abandoned algorithm after new compute or data changes and reframe it
- Probe robustness by systematically violating evaluation assumptions (distributional, adversarial, scale)
- Decompose a monolithic pipeline to identify the true bottleneck and a minimal competitive baseline
- Resolve a trade‑off (e.g., privacy vs utility) by characterizing whether the gap is fundamental or an artifact
FAQ
You can use a single framework for focused exploration, but combining 2–3 complementary lenses typically yields stronger, more defensible ideas.
When should I stop ideating and start executing?
Stop when an idea passes the two‑sentence test, names stakeholders, and exposes at least one concrete, testable assumption you can validate quickly.