Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.
CommunityRun safe pilot experiments with confirmations.
Education & Research#research automation#results reporting#baseline comparison#YAML spec#GPU training#remote SSH#pilot experimentation
Authorduany049
Version1.0.0
Installs0
System Documentation
What problem does it solve?
It streamlines turning a Pilot Spec YAML into runnable pilot code and quickly getting diagnostic results without directly altering wiki pages or producing a verdict.
Core Features & Use Cases
- Pilot code generation from spec: Creates training/evaluation code under experiments/pilot/code/{slug}/ using the pilot_spec fields (setup, metrics, baseline, success criterion, hypothesis, and approach sketch).
- Local or remote execution with guardrails: Supports local GPU runs and remote SSH execution via tools/remote.py, including connectivity and GPU availability checks.
- Human-in-the-loop safety: Requires explicit user confirmation before launching runs and instructs manual inspection of code/configurations; does not judge pass/fail and does not modify wiki content.
Quick Start
Ask the AI to run the pilot for idea slug demo-idea in local mode and return the raw results for review.
Dependency Matrix
Required Modules
None requiredComponents
assets
💻 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: Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail. Download link: https://github.com/duany049/skill-offline-auto-opti/archive/main.zip#pilot-experiment-execution-read-pilot-spec-yaml-write-pilot-code-run-experiment-confirm-with-the-user-before-operation-and-require-the-applicant-to-conduct-manual-inspection-return-results-called-by-ideate-phase-5-does-not-modify-wiki-pages-or-judge-pass-fail 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.