# Clawpathy Autoresearch

Eval-driven skill tuning.

- Repository: https://github.com/ClawBio/ClawBio (1172 stars)
- Category: AI and agents
- License: MIT
- Last updated: 2026-11-01
- Runs scripts: yes, review before use

## Install

Claude Code: `npx skills add ClawBio/ClawBio --skill clawpathy-autoresearch -g -a claude-code -y`
Codex: `npx skills add ClawBio/ClawBio --skill clawpathy-autoresearch -g -a codex -y`
Cursor: `npx skills add ClawBio/ClawBio --skill clawpathy-autoresearch -g -a cursor -y`
Gemini CLI: `npx skills add ClawBio/ClawBio --skill clawpathy-autoresearch -g -a gemini-cli -y`
Claude Code plugin marketplace: `/plugin marketplace add ClawBio/ClawBio`

## Description

Eval-driven skill tuning. Given a task and an LLM-judge rubric, iteratively rewrites a SKILL.md until a downstream executor agent performs well against the judge. Low-code: all evaluation is LLM-as-judge, not deterministic Python.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# clawpathy-autoresearch

Eval-driven skill development. The system iteratively rewrites a `SKILL.md`
so a downstream executor agent performs better at a task class, as judged
by an LLM against a paper/task-specific rubric.

## Core idea

```
  propose (sonnet)  →  execute (sonnet, shell)  →  judge (opus, rubric)
       ↑                                                       │
       └──────── feedback: verdict + recommended edits ────────┘
```

- **Proposer** rewrites SKILL.md based on the last judge verdict.
- **Executor** runs the new SKILL.md end-to-end inside a workspace.
- **Judge** scores methodology (primary) and outputs (secondary) against
  a per-task rubric. Lower is better; 0 = perfect.
- Keep the new SKILL.md only if it strictly beats the best score; else
  revert. Stop on target_score or on `early_stop_n` consecutive regressions.

## You are the orchestrator

You (the agent reading this) don't run the loop yourself. You dispatch
subagents to build the workspace, then hand off to the Python loop.

### Phase 1, Scout

Dispatch a subagent with `prompts/scout.md` to research the paper/task.
Report key findings to the user in a few lines.

### Phase 2, Scope (you + user)

Have a conversation. Ask ONE question at a time, multiple-choice where
helpful. Agree on:
- what to reproduce / what success looks like
- which data sources are in-bounds
- what methodology expectations belong in the rubric
- iteration budget and target_score (if any)

Present a summary and get approval.

### Phase 3, Build

Dispatch a builder subagent with `prompts/builder.md` and the agreed
scope. It writes:
- `task.json`
- `rubric.md`, **the authoritative scoring rubric for the LLM judge**
- `reference/` (optional; judge-only)
- `skill/SKILL.md`, seed

Validate:
```python
from skills.clawpathy_autoresearch import validate_workspace
print(validate_workspace(Path("WORKSPACE")))  # [] means valid
```

### Phase 4, Loop

The rest: https://github.com/ClawBio/ClawBio/blob/HEAD/skills/clawpathy-autoresearch/SKILL.md