Trio for Pi is a plugin for the pi.dev platform that implements a single-session planning, execution, and review workflow using different AI models for each phase.

Stars

35

7-day growth

No data

Forks

0

Open issues

0

License

MIT

Last updated

2026-07-13

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It introduces a structured three-phase workflow (planner → executor → reviewer) within a single session, enabling multi-model collaboration and iterative refinement on the pi.dev platform.

Who it is for

  • Developers using pi.dev for AI-assisted coding
  • Researchers exploring multi-step AI agent workflows
  • Engineers seeking automated code review within a single conversation
  • Users wanting configurable role-specific models and prompts

Use cases

  • Automated code generation with built-in review iterations
  • Multi-step reasoning tasks requiring different model strengths
  • Content creation where planning and revision are separated
  • Rapid prototyping of complex prompts with feedback loops

Strengths

  • Shared conversation and working tree across all three phases
  • Easy installation via `pi install` and simple setup with `/trio setup`
  • Configurable models, system prompts, and review round limits per role
  • Supports both global and project-local configuration overrides

Considerations

  • Requires the pi.dev platform as a dependency
  • Limited documentation beyond basic command descriptions
  • Single-session scope — no persistence or multi-turn support beyond one workflow

README quick start

Trio for Pi

A single-session planner → executor → reviewer workflow for pi. Trio keeps one shared conversation and working tree while switching models between phases.

Install

pi install git:github.com/jnsahaj/trio

Get Started

pi
/trio setup

Configuration is saved to ~/.pi/agent/trio.json. Run /trio setup to choose again, or edit that file to change role models and optional role settings.

Roles accept optional thinkingLevel and systemPrompt fields. The optional top-level maxReviewRounds field sets a review-round cap. A standalone configuration includes a provider and model for all three roles. A project-local .pi/trio.json can override fields from a complete global configuration.

Other commands: /trio status, /trio config, /trio stop.

Description

A planner, executor, and reviewer workflow for Pi

Related repositories

Similar projects matched by category, topics, and programming language.

TanStack
Featured
TanStack GitHub avatar

router

TanStack Router is a type-safe, data-driven React router with built-in caching, prefetching, and nested layouts, while TanStack Start extends it into a full-stack SSR framework.

Web DevelopmentFrontend Frameworks
14,861
vercel-labs
Featured
vercel-labs GitHub avatar

scriptc

scriptc compiles ordinary TypeScript into small, fast native executables without needing Node.js, V8, or any JavaScript runtime in the binary.

Developer ToolsCode Quality & Build
1,985
Jakubantalik
Featured
Jakubantalik GitHub avatar

thinking-orbs

A React component library that renders six hand-tuned animated thought orb loading indicators on a plain 2D canvas, with two purpose-tuned sizes and automatic theme detection for AI and agent UIs.

AI & Machine LearningAI Agents
1,191