Commonly provides a chat interface that lets developers collaborate with multiple AI agents—such as native agents, OpenClaw, Codex, and Claude—within a single roster. Each agent retains a persistent identity and memory tied to the project, so when runtimes are swapped or new agents are added, they remember prior interactions and skills. The platform supports both hosted agents for zero‑setup use and self‑hosted agents that can be imported with their existing memory.
Funding
Funding not disclosed
Founders
Product
Problem
Developers often need to switch between different AI coding assistants or integrate custom agents, but each tool maintains its own isolated context, requiring repeated setup and re‑briefing. This fragmentation hampers seamless collaboration and slows down tasks such as specification writing, pull‑request generation, and code review.
Solution
Commonly offers an open‑source platform that unifies AI agents into a single, persistent roster within a development project. Each agent retains a distinct identity, skill profile, and long‑term memory that persists across runtime swaps, allowing developers to add, replace, or connect agents without losing context. Hosted agents are available for immediate, zero‑setup use, while teams can also attach their own agents or any HTTP‑based endpoint. The shared memory lives with the project, enabling all agents and human teammates to read and write the same context, which streamlines collaborative workflows such as spec creation, PR drafting, and code review.
Target Audience
Primary users are software development teams and individual developers who integrate AI assistants into their coding workflow and need a consistent, collaborative environment for code generation and review.
Features
- Unified roster (“pods”) that hosts multiple agents (e.g., native, OpenClaw, Codex, Claude Code) side by side
- Persistent agent identity and memory that survive runtime changes or swaps
- Hosted zero‑setup agents plus support for connecting custom or self‑hosted agents via HTTP webhooks
- Direct 1:1 messaging with any agent in the roster
- Skill and specialty profiles displayed on each agent’s profile for easy selection
- Memory shared at the project level, eliminating the need for manual context transfer when adding new teammates or agents