pi-subagents

Overview of pi-subagents

What it is

A pi extension that brings Claude Code-style autonomous sub-agents and workflow orchestration to pi. It allows spawning specialized agents that run in isolated sessions, each with its own tools, system prompt, model, and thinking level. Agents can run in the background or block, be steered mid-run, resume completed sessions, and use custom agent types. It also includes a deterministic JavaScript workflow script tool (SubagentWorkflow) that supports agent(), parallel(), pipeline() and is compatible with Claude Code's Workflow tool.

Key features

• Parallel background agents with automatic queuing (configurable concurrency limit, default 10)
• Live widget UI showing animated spinners, tool activity, token counts, and colored status icons
• FleetView: a navigable list of main and running subagents below the editor
• Custom agent types defined in .pi/agents/.md or globally with YAML frontmatter
• Mid-run steering and session resume (preserves full conversation context)
• Scripted workflows using JavaScript with agent(), parallel(), pipeline(), phase(), log()
• Package-shipped agents: a pi package can declare its own subagents via package.json

Use cases

• Orchestrating multiple specialized AI agents in parallel for complex tasks
• Running background research agents while continuing the main conversation
• Automating multi-step workflows with deterministic scripts
• Creating reusable agent types for specific roles (e.g., explore, audit, test)
• Integrating with existing Claude Code workflows and agent definitions

Getting started

See the repository README for installation and usage instructions.

License and pricing

License: MIT. Pricing: open source.

Based on the project's public README and repository metadata.