Releases·Tool Release
Codex exposes local endpoints for all plugins without extra auth, enabling use from Pi or any harness
A developer released @wileai/pi-codex-connector and @wileai/codex-connectors-mcp, letting any model or agent harness use Codex plugins via locally exposed server endpoints requiring no extra auth. Pi users install the connector with one command; other harnesses launch an MCP server via npx. The release responds to the discovery that Codex exposes plugin endpoints without additional authentication.
2 engagement1 sourceFri, Oct 2, 2026, 02:49 AM
AI Overview
- Codex now exposes local server endpoints for all plugins without extra auth, usable from any model or harness
- Developer releases @wileai/pi-codex-connector to install Codex plugins inside Pi via one shared auth
- Standalone @wileai/codex-connectors-mcp npm package lets any harness access Codex plugins via npx
Related
Releases·Tool Release
resume-from lets developers hand off coding sessions across Pi, Claude Code, and Codex
The maintainer of `resume-from` released a local session handoff tool that lets developers continue coding sessions across Pi, Claude Code, and Codex without sharing raw vendor-specific session files. It addresses the problem of rebuilding task context when switching agents or hitting usage limits.
The maintainer of `resume-from` released a local session handoff tool that lets developers continue coding sessions across Pi, Claude Code, and Codex without sharing raw vendor-specific session files. It addresses the problem of rebuilding task context when switching agents or hitting usage limits.
2 engagement1 sourceredditFri, Aug 7, 2026, 03:04 PM
Products·launched
Open-source interoperability layer for AI agents
A2A Adapter and Hybro are open-source projects that enable different AI agents and harnesses to work together. They turn existing agents into interoperable A2A agents and coordinate their collaboration, solving the problem of agent fragmentation.
A2A Adapter and Hybro are open-source projects that enable different AI agents and harnesses to work together. They turn existing agents into interoperable A2A agents and coordinate their collaboration, solving the problem of agent fragmentation.
10 engagement1 sourceredditThu, Aug 20, 2026, 04:21 PM
Products·launched
mecha: terminal multi-agent orchestrator with coordinator and specialists
mecha is a single-binary terminal tool that lets a coordinator agent break down user requests and dispatch tasks to specialist agents (architect, coder, tester, reviewer) running in separate terminal panes. It works with existing agent CLIs like Claude Code, Codex, Gemini, CodeBuddy, and Pi, and uses agent hook events for task state, avoiding polling. It solves the problem of coordinating multiple AI agents for complex tasks, providing visibility and control for developers.
mecha is a single-binary terminal tool that lets a coordinator agent break down user requests and dispatch tasks to specialist agents (architect, coder, tester, reviewer) running in separate terminal panes. It works with existing agent CLIs like Claude Code, Codex, Gemini, CodeBuddy, and Pi, and uses agent hook events for task state, avoiding polling. It solves the problem of coordinating multiple AI agents for complex tasks, providing visibility and control for developers.
10 engagement1 sourceredditSun, Aug 2, 2026, 01:22 AM
Releases·Tool Release
Developer releases open-source pack of 54 reusable coding-agent skills
A developer released an open-source collection of 54 reusable skills for coding agents including Claude Code, Codex, OpenCode, and Pi. The pack aims to standardize common agent workflows, making them portable across multiple tools.
A developer released an open-source collection of 54 reusable skills for coding agents including Claude Code, Codex, OpenCode, and Pi. The pack aims to standardize common agent workflows, making them portable across multiple tools.
1 engagement1 sourceredditFri, Aug 14, 2026, 04:06 AM
Products·building
Nimrod: cross-agent work memory for Claude Code, Codex, OpenCode and Pi
Nimrod is a local tool that gives coding agents shared work memory across sessions and across agents. It reads each agent's own on-disk session stores (JSONL or SQLite, read-only), extracts useful context like requests, files touched, commands run, and session outcomes, so switching between Claude Code, Codex, OpenCode, or Pi doesn't reset project context. It runs locally with no API keys, solving the pain of re-explaining the same project to different agents.
Nimrod is a local tool that gives coding agents shared work memory across sessions and across agents. It reads each agent's own on-disk session stores (JSONL or SQLite, read-only), extracts useful context like requests, files touched, commands run, and session outcomes, so switching between Claude Code, Codex, OpenCode, or Pi doesn't reset project context. It runs locally with no API keys, solving the pain of re-explaining the same project to different agents.
5 engagement1 sourceredditSun, Sep 20, 2026, 02:48 PM