CtrlK
BlogDocsLog inGet started
Tessl Logo

brainstorm

Run interactive brainstorming across verifiers tasksets, evaluations, GEPA, and RL training. Use when the user wants ideation, literature scanning, concept teaching, roadmap planning, or research program design grounded in local CLI sources, verifiers, and RL trainer code.

68

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Brainstorm

Goal

Run an interactive idea session with the user to turn their idea into a concrete taskset outline.

Interaction style

  • Find out what the goal of the user is, what their budget is, and ultimately what they want to achieve.
  • Ask a lot of clarifying questions and outline the concrete steps before you start implementing.
  • When something is ambiguous, do research first and present the user with different options to choose from.
  • Don't overload the user with terminology, but guide them in the right direction. Assume a knowledgeable user, but not an in-depth expert at every topic.

Discovery workflow

  1. Find out what the user wants:
  • Is it running existing evaluations or training with common tasksets? If so, use the Environment Hub to see whether those tasksets already exist.
  • Is it optimizing an existing workflow? In this case, prompt optimization with GEPA is the right tool.
  • Else: For training and evaluations, building a taskset together is the right call.
  1. After you have found this out, look at the other skills and choose the appropriate one to proceed on the technical level, while still keeping the interactive session with the user alive.
Repository
PrimeIntellect-ai/verifiers
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.