Skip to content
Get packs

Experiment Bridge

Turns an experiment plan into implemented code, deployed runs, and initial results.

Source Download skill
Repository
wanshuiyin/Auto-claude-code-research-in-sleep
Stars
18,482
Licence
MIT
Files
1 file

What it does

This skill is for research workflows. After you have an experiment plan (an EXPERIMENT_PLAN.md file), the agent reads it, writes the experiment code (training, evaluation, data loading, baselines), optionally has a second AI reviewer check the code for logic bugs, and then deploys the runs to a GPU and collects first results. It is a bridge between idea generation and an automated review loop.

Great for

  • Researchers who already have a written experiment plan and want the agent to implement and launch it
  • Students or engineers running machine-learning experiments who want code scaffolding with argparse, seeds, and JSON/CSV result outputs

Skip it if

  • Projects without an experiment plan file: the skill expects planning documents from an earlier workflow and otherwise asks what to implement
  • Running experiments on shared or paid GPU infrastructure without review: with AUTO_DEPLOY enabled, the agent deploys code automatically, which can spend GPU hours and money on buggy code

Try asking

  • “Implement the experiments in refine-logs/EXPERIMENT_PLAN.md and deploy the sanity run first”
  • “Turn my experiment plan into code and collect initial results for the review loop”
  • “Deploy the plan to my GPU server over SSH and report the first metrics”
Step-by-step guide: how to use Experiment BridgeInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    experiment-bridge.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip experiment-bridge.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip experiment-bridge.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Review the generated code before deployment by setting AUTO_DEPLOY to false, especially for expensive runs
  • Make sure the experiment plan and FINAL_PROPOSAL.md are up to date so the implementation matches the intended method
  • Confirm which backend (local, SSH, Vast.ai, Modal) you want, since the skill is designed not to silently fall back to local execution

What's inside

  • SKILL.md

Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.