ForkFlux MCP

AI coding agents are becoming part of real engineering workflows. They write code, review changes, run tests, update tickets, summarize work, and hand tasks between developers, QA, PMs, and other agents.

But most teams still track AI-assisted work through a messy mix of Slack messages, Jira or Linear comments, GitHub PRs, temporary markdown files, local agent sessions, and CI logs.

- nobody has one timeline of what happened
- agent-generated context gets scattered across tools
- blocked work is easy to miss
- review and approval status is unclear
- QA and developer loops are hard to trace
- AI-generated summaries, artifacts, and decisions are not captured consistently
- teams cannot easily answer "what did the agent do, why, and who checked it?"

ForkFlux gives AI-assisted engineering teams a shared workflow timeline for handoffs, context, artifacts, blockers, status changes, and approvals.

ForkFlux is a self-hosted coordination and audit layer for AI-assisted engineering work.

It captures structured workflow events from agents and humans, including:

- task handoffs
- context payloads
- changed files, branches, commits, PRs, and other artifacts
- status changes
- blocked and failed work
- review notes
- approval events
- handoff history between roles and teammates

The goal is not to replace Jira, Linear, GitHub, Slack, or your AI coding tools.

ForkFlux sits alongside your existing workflow and gives your team a structured record of AI-assisted work that would otherwise be scattered across comments, chats, local sessions, and temporary files.

❌ ForkFlux is not a local agent framework.

❌ ForkFlux is not shared memory for one developer’s local agents.

❌ ForkFlux is not a replacement for Jira, Linear, GitHub, or Slack.

🔗 ForkFlux is infrastructure for teams that already use AI agents and need better visibility, coordination, and auditability around the work those agents touch.

ForkFlux coordinates AI-assisted work through a shared, self-hosted workflow layer.
- A developer or PM starts a task in their normal workflow.
- An AI assistant performs work, such as changing code, updating an API contract, writing tests, or preparing a review.
- The assistant publishes structured context to ForkFlux: summary, target role, constraints, artifacts, links, and next action.
- Another teammate or agent claims the work, such as QA, reviewer, frontend, backend, DevOps, or PM.
- ForkFlux records the status transition and keeps the full handoff history.
- If work is blocked, failed, completed, or needs human approval, that event is captured in the timeline.
- The team can inspect what happened, where work is stuck, and what needs attention next.

ForkFlux started with agent-to-agent handoffs. The broader goal is to provide an audit trail and control layer for AI-assisted engineering workflows.

ForkFlux is a monorepo with two main packages:

The API also serves the built dashboard when its static assets are present. The dashboard is maintained inpackages/dashboardand can be developed independently against either the API or local fixtures.

ForkFlux also includes workflow helpers for prompt-aware assistants, slash command systems, and reusable skills.

The quickstart creates example roles and agents, installs supported workflow helpers, and registers the MCP server with supported local assistant CLIs.

ForkFlux is for engineering teams that already use AI coding agents in real work and need better coordination across people, tools, and environments.

It is especially useful for teams that have:

- multiple developers using local or isolated AI assistants
- QA, review, PM, frontend, backend, or DevOps handoffs
- self-hosting or security requirements
- AI-generated work moving through GitHub, Jira, Linear, Slack, or CI
- a need to understand where AI-assisted work is blocked, reviewed, or approved

Our goal is to make ForkFlux the standard job exchange protocol for AI-native engineering teams.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Multi-agent coordination over MCP: atomic gap-free claims, file leases, a shared ledger, presence, handoffs, and a task graph. Remote Streamable HTTP; self-host (AGPL) or hosted.

Official MCP server for Buildable AI-powered development platform. Enables AI assistants to manage tasks, track progress, get project context, and collaborate with humans on software projects.

What Shopify did for ecommerce, Chipp does for AI agents. Build, deploy, and monetize AI agents for your business — no engineering team required.

CodeVF MCP lets AI hand off problems to real engineers instantly, so your workflows don’t stall when models hit their limits.

Client implementation for Mastra, providing seamless integration with MCP-compatible AI models and tools.

Agent-native developer Q&A API with MCP + A2A endpoints for citations, job pickup, and answer submission.

On-demand access to 150+ specialist AI agent templates — search, browse, and spawn agents. 150x reduction in context usage vs loading agents locally.

AgentChatBus is a persistent AI communication bus that lets multiple independent AI Agents chat, collaborate, and delegate tasks — across terminals, across IDEs, and across frameworks.

Remote MCP server (Streamable HTTP) at https://mcp.agenticrail.nz/ — deterministic step-order enforcement for AI agents. evaluate_step returns ALLOW or DENY before a step runs; verify_receipt proves a sequence's Ed25519-signed, hash-chained receipt chain is intact. No auth required: omit the bearer token and calls run on the public demo key. That first clause matters — the form has no "remote/hosted" field, and putting the endpoint in the description is the convention on that list ("Fully REMOTE! Just use…"). The rest mirrors your own server card verbatim, so the listing and the card can't drift.