Pi Coding Agent

Pi Durable persists in-flight tools, application state, and subagents so long-running agents can suspend and resume across environments. Demonstrations show checkpointed memory, approvals, and desktop control across devices.

Raunak demonstrates a Pi Durable agent that resumes after redeploying itself. Its Durable Object setup stores memory and chat history in SQLite, preserving both across deployments.

Mario Zechner demonstrates a native Android app built on his phone with Pi Durable in two days. The prototype also adds paged transcript loading, but he describes it as a demo rather than a product.

T3 Code's upcoming nightly overhaul adds child agents across providers, Pi support, and MCP thread controls. Its creator warns of instability as model switching, queueing, and automatic resumption are introduced.

Pi 1.0 introduces Pi Durable, with SQLite storage and concurrent sessions shared across multiple clients. Published examples describe replay-safe tasks and background subagents.

Rivet's Pi integration keeps the agent loop, sessions, and credentials in the backend while sandboxes provide tools. Rivet reports durable execution, shared control, and SQLite persistence alongside the 0.82 MB per-session figure.

Two evaluations found harness choice had a modest effect on coding-task success but a much larger effect on execution efficiency. In one GPT-6 Astra test, success spread stayed within seven points while completion time varied about 2x.

Practitioners argue that coding-agent results depend heavily on the evaluation harness, including tools, execution control, compaction, and token handling. They propose stable common harnesses rather than vendor-specific setups.

Community tests report Qwen 3.8 27B handling coding, OCR, and long-context workloads locally. One vLLM setup reached 3,200 tokens per minute at 262K context on two RTX 3090s without NVLink.

Local users report that Qwen 3.8 27B agent behavior changes substantially with the harness, quantization, context settings, and hardware. One 3-bit MacBook Air run at 57K context took 63 hours.

AI SDK canary added HarnessAgent, a unified abstraction that runs Claude Code, Codex, and Pi in sandboxed sessions with AI SDK-compatible streams. One integration can now target multiple agent harnesses without separate model-specific plumbing.

Three independent Pi builders shipped a goal runner, contract-style subagent acceptance gates, and a new Lovely Dev Tools extension in the same window. That gives Pi users more deterministic long-running loops and cleaner local tool interfaces without starting from an empty harness.

Pi v0.75.5 now shows only the read line in collapsed tool cards while keeping the full inspected range behind Ctrl+O. That matters because long read outputs were obscuring edits and steering signals in collaborative coding sessions.

Pi raised its minimum Node version from 20 to 22.19.0, then shipped a follow-up after Undici-related changes in Node 26 caused Copilot and Codex login failures. This matters because agent CLIs built on Node and Undici can hard-fail on auth or install paths after runtime upgrades.

Builders shipped pi-treebase, a Miko voice mode for pi-listens, devrage support, and a Japanese OpenCode Go guide after the first Pi extension burst. The releases arrive as Pi’s provider abstraction gets stress-tested by OpenClaw-scale multi-provider use.

Independent Pi builders shipped a voice layer, a kanban and observability dashboard, a Codex-conversion tool with `apply_patch`, and smaller UI extensions in the same window. The burst matters because it turns Pi from a single coding agent into a real local-first extension ecosystem with voice, review, and workflow primitives.