pi-harness
The operational superset of Pi Coding Agent — everything Pi, plus observability, governance, recovery, evaluation and multi-agent orchestration. Pi Coding Agent
What is pi-harness?
The operational superset of Pi Coding Agent — everything Pi, plus observability, governance, recovery, evaluation and multi-agent orchestration. Pi Coding Agent 的运维超集——完整继承 Pi 的核心能力,并进一步扩展可观测性、治理、故障恢复、评估和多 Agent 编排。
Overview of pi-harness
What it is
Pi-Harness is the operational superset of Pi Coding Agent. It brings Pi Coding Agent into a desktop workspace for macOS, Windows, and Linux. Users can manage conversations, projects, files, Git, models, and Skills in one place, with tools to inspect runs, control budgets, recover work, and coordinate agent teams. Native Pi runs the agent, and Pi sessions, configuration, tools, Skills, and extensions remain compatible with Pi.
Key features
- Workspace, files, and Git: start standalone chats or open projects, browse and search files, preview media, edit text with syntax highlighting, and use Git diffs, history, commits, and worktrees.
- Providers and models: configure Pi-compatible providers with presets or custom endpoints, manage credentials, discover models, test connections, and choose chat models.
- Skills, extensions, packages, and MCP: browse built-in and local Skills, edit them, manage Pi packages and extensions, connect MCP through supported packages, and install optional add-ons.
- Harness: observe, control, and recover: inspect Runs, Trace, Replay, timelines, and tool activity; set tool policies and per-run budgets; create checkpoints; resume, fork, or retry work; use Evaluation, Diagnostics, and regression baselines.
- Agent teams and task orchestration: coordinate Pi agents with reusable teams and templates; assign tasks with dependencies and priorities; schedule manually, sequentially, or in parallel; use review gates and worktree isolation.
- Local conversation history: browse Claude Code, Codex, Cursor, Gemini CLI, OpenCode, and Pi history together; search, filter by source, project, or date range; view original conversations in read-only mode and continue in new Pi chats.
- Project architecture: explore modules, dependencies, entrypoints, and source references; filter by package, focus on a module, save analysis snapshots, and export JSON, offline HTML, or PNG.
Use cases
Developers using Pi Coding Agent who need observability, governance, recovery, evaluation, and multi-agent orchestration. Teams coordinating multiple agents on complex tasks. Managing projects with Git, files, and conversation history in a unified desktop workspace.
Getting started
See the repository README for installation instructions.
License and pricing
License: AGPL-3.0. Pricing: open source (free).
Based on the project's public README and repository metadata.