{"slug": "proliferate-unifies-claude-code-codex", "title": "Proliferate unifies Claude Code, Codex", "summary": "Proliferate, an open-source, self-hostable AI IDE from YC S25, unifies Claude Code, OpenAI Codex, OpenCode, Cursor, and Grok in a single interface, allowing inter-agent delegation and reusable workflows. The tool, available on GitHub under AGPL-3.0, lets users run multiple agents with native inference paths and human approval gates, with a typical PR workflow running in about 12 minutes. It currently lacks Windows support, requires ~24 GB RAM for three concurrent agents, and has a minimal agent config UI.", "body_md": "# Proliferate unifies Claude Code, Codex\n\n[Claude Code](/en/tags/claude%20code/)refactor the implementation, and Grok review the PR — all without leaving a single terminal — I stopped paying for three separate subscriptions. That was two weeks ago. I've been running Proliferate on a home server since, and it's the first tool that actually delivers on the \"one interface for every agent\" promise without feeling like a compromised wrapper.\n\nProliferate (GitHub: proliferate-ai/proliferate) is an open-source, self-hostable AI IDE from YC S25 that lets you drive Claude Code, OpenAI Codex, OpenCode, [Cursor](/en/tags/cursor/), and Grok from the same UI. Each agent keeps its native inference path — Bedrock, Azure, local vLLM, whatever you configure — so you're not locked into a lowest-common-denominator abstraction layer. The AGPL-3.0 license means you can audit, fork, or run it behind your own VPN without asking permission.\n\n## Inter-agent delegation that works\n\nThe feature that changed my workflow: a parent agent can spawn subagents and pass context between them. My current loop:\n\n1. **Fable** (my planning agent) breaks down a feature spec into discrete tasks\n\n2. **Codex** implements the core logic with its computer-use tooling\n\n3. **OpenCode** with a local Qwen2.5-Coder model reviews the diff for style and security issues\n\n4. ** Claude Code** writes the integration tests and updates documentation\n\nAll four sessions share a persistent workspace. The handoff is explicit — you see each agent's output, approve or reject, then pass the baton. No hidden state, no \"trust me bro\" black boxes.\n\n```\n# Example workflow snippet from my code-review pipeline\nsteps:\n  - agent: codex\n    task: \"implement user-auth module per SPEC.md\"\n    model: gpt-4o\n  - agent: opencode\n    task: \"review diff for OWASP top 10 violations\"\n    model: qwen2.5-coder:32b\n    approval_gate: true\n  - agent: claude-code\n    task: \"write integration tests for auth flow\"\n    model: claude-3-5-sonnet\n```\n\n## Workflows = reusable automation, not chat history\n\nWorkflows are where Proliferate separates from \"just another chat UI.\" You define a chain of agent sessions with human approval gates, document passing, and per-step model selection. I built one for our PR process: spec ingestion → implementation → security review → test generation → changelog draft → final human sign-off. Runs in about 12 minutes end-to-end. The YAML is version-controlled alongside the repo, so the automation evolves with the codebase.\n\n## Rough edges worth knowing\n\n**No Windows support yet**— Docker on WSL2 works but expect path-mapping friction** Agent config UI is minimal**— you'll edit JSON for Bedrock/Azure credentials** Streaming output occasionally desyncs**between the frontend and long-running Codex sessions** Resource usage**— running three agents concurrently needs ~24 GB RAM on my box\n\nThe team ships weekly. Rough spots get patched fast.\n\nIf you're evaluating AI workflow tooling for a team that refuses vendor lock-in, spin up the Docker Compose stack and throw a real task at it. The self-hosted story is genuine — no telemetry, no external calls unless you configure them.\n\n[Built a shared brain for AI agents after markdown file chaos 1h ago](/en/news/7235/)\n\n[Built a schedule-aware PM copilot that actually respects 1d ago](/en/news/7077/)\n\n[Built a remote control for the AI agents living on my garage 2d ago](/en/news/6905/)\n\n[Does anyone actually care about llms. 3d ago](/en/news/6847/)\n\n[AI is making every side project feel like meaningless slop 3d ago](/en/news/6755/)\n\n[HarnessRouter makes it way easier to switch between Claude Code 4d ago](/en/news/6726/)\n\n[Next Why the current alignment toolkit still leaks — and what →](/en/news/7239/)", "url": "https://wpnews.pro/news/proliferate-unifies-claude-code-codex", "canonical_source": "https://promptcube3.com/en/news/7241/", "published_at": "2026-08-22 00:24:45+00:00", "updated_at": "2026-08-22 00:42:22.620853+00:00", "lang": "en", "topics": ["ai-tools", "developer-tools", "ai-agents", "generative-ai"], "entities": ["Proliferate", "YC S25", "Claude Code", "OpenAI Codex", "OpenCode", "Cursor", "Grok", "GitHub"], "alternates": {"html": "https://wpnews.pro/news/proliferate-unifies-claude-code-codex", "markdown": "https://wpnews.pro/news/proliferate-unifies-claude-code-codex.md", "text": "https://wpnews.pro/news/proliferate-unifies-claude-code-codex.txt", "jsonld": "https://wpnews.pro/news/proliferate-unifies-claude-code-codex.jsonld"}}