plan-experiment
OfficialTurn a research idea into a runnable plan.
Education & Research#research planning#mechanistic interpretability#session workflow#causal abstraction#hydra configs#experiment DAG#cache reuse
Authorgoodfire-ai
Version1.0.0
Installs0
System Documentation
What problem does it solve?
This Skill converts a fuzzy research goal into a complete experiment plan, defining the causal task setup and the full analysis dependency DAG so downstream execution and interpretation can run reliably.
Core Features & Use Cases
- Objective crystallization: Produces a structured
RESEARCH_OBJECTIVE.mdwith an evidence-oriented objective, optional hypotheses, and success criteria. - Dependency-DAG planning: Builds
PLAN.mdwith the analysis chain (baseline → locate → subspace → activation/manifold geometry checks, etc.), including per-node I/O contracts and measurable pre-flight gates. - Runner configuration strategy: Specifies how to materialize runner configs for
/run-experiment, including sweep and cache-reuse considerations to reduce redundant compute.
Quick Start
Start a new research investigation by running plan-experiment to generate the session plan artifacts that you will execute next with /run-experiment.
Dependency Matrix
Required Modules
None requiredComponents
Standard package💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: plan-experiment Download link: https://github.com/goodfire-ai/causalab/archive/main.zip#plan-experiment Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.