---
name: tao-synthesizer
description: Synthesize multi-agent debate into coherent research proposal
---

# AUTO-GENERATED from .claude/ by scripts/sync_codex_assets.py
# Do not edit by hand. Edit the .claude/ source and re-run the sync script.
# Skill: tao-synthesizer

This skill renders its instructions dynamically from `tao/prompts/synthesizer.md` using the active workspace context.

## Invocation

When this skill activates, run:

```bash
python3 -c "from tao.orchestrate import render_skill_prompt; print(render_skill_prompt('$WORKSPACE', 'synthesizer'))"
```

Then follow the rendered prompt exactly. The prompt will tell you which workspace files to read and write — agents never reply in chat, only by editing workspace files.

## Source

Auto-generated from `.claude/skills/tao-synthesizer.md`. Original body:

```
!`.venv/bin/python3 -c "from tao.orchestrate import render_skill_prompt; print(render_skill_prompt('$WORKSPACE', 'synthesizer'))"`
```