Bring your own agent
Claude Code, Claude CLI, OpenAI Codex, Pi, Cursor, and Grok. Hand a live session over without losing the transcript.
Agentic Coding Environment
A workspace where your projects live, your agent works, and your apps get built and shipped.
You and your coding agent of choice.
The full dev cycle, one continuous flow
Choose your collaborator
Pick the agent per session. Switch or hand a running session over at any time — the same pm7Code session keeps its identity and transcript.
Anthropic, via the Claude Agent SDK
Anthropic's Claude Code CLI
OpenAI's Codex
Multi-provider, defaults to Google Gemini
Cursor's agent over ACP
xAI's Grok over ACP
Optional Orchestrator mode lets one agent delegate sub-tasks to the others. pm7Code carries its own protocol into every agent, keeping behavior consistent without depending on a global config file.
The workspace
Projects, agents, builds, and deploys — all in the same place, all visible at the same time.
Claude Code, Claude CLI, OpenAI Codex, Pi, Cursor, and Grok. Hand a live session over without losing the transcript.
Sessions, prompts, and history live with the project — not with whichever agent thread happened to run last.
Run builds, ship deploys, and watch logs without leaving the workspace or losing the prompt that started the work.
Prompts, shelves, and history give you structured context you do not have to rebuild for every task.
A readable history of agent runs, questions, tool calls, results, and summaries that you can search and replay.
System, Group, Project, and Session prompts inherit in a clear order, with visibility into what reached the agent.
Answer structured agent questions inline, with optional free text. The agent continues without guessing.
Ask a designated buddy agent for a second opinion and see both recommendations side by side.
Commit and push continue as a visible background job while the active session keeps working.
Preview the deploy you just shipped in an optional browser panel beside the transcript.
Independent summary slots distill long runs so you can scan what an agent is doing at a glance.
Project and global memory stay agent-independent. The agent proposes what matters; pm7Code decides what persists.
Keep moving forward
Everything between “I have an idea” and “it’s shipped” stays in the same place.
Open documentation