Skip to main content
Glama

Website Image License: MIT Build

Localized · Self-hosted · No paid tiers · No enterprise gates · No vendor lock-in

Roadmap · Governance · Contributing · Code of Conduct


What is Qadam Flow?

Qadam Flow is an open-source workflow automation platform — connect your apps and services with visual flows, no code required. It is built on the Activepieces MIT core, with all enterprise-licensed code removed, so every feature is free and open.

It is the first project of AI Qadam Build — open infrastructure that Central Asia's AI community owns and runs itself.

Related MCP server: mcpgate

Why it exists

The region runs its automation on foreign SaaS — billed in dollars, hosted abroad, in languages that aren't ours, with the features that matter (SSO, RBAC, audit logs) locked behind enterprise paywalls. Qadam Flow is the region's own: localized, self-hosted, data kept local, gates removed. The point isn't "a free Zapier" — it's independence.

Status

Early, and honest about it. The Activepieces MIT core is forked and cleaned. Localization and regional integrations are the current focus; enterprise features (SSO / RBAC / audit) are built as real deployments need them, not on a fixed schedule — see the Roadmap. Maintained by AI Qadam — founder-stewarded for now, community-owned by design (DCO, no copyright assignment).

Quick start

Requires Docker. One line, no clone, no build:

curl -fsSL https://flow.aiqadam.org/run.sh | sh

This drops a fresh ./qadam-flow/ next to your shell, pulls ghcr.io/aiqadam/qadam-flow:latest, generates random secrets in .env, brings up postgres + redis + app + workers, and prints the URL when the API is ready. Works on macOS (Docker Desktop), Linux (dockerd), and Windows via WSL2.

Then open http://localhost:8080/sign-up — the first signup owns the platform.

Env knobs (export before piping into sh):

Variable

Default

Purpose

QADAM_FLOW_DIR

./qadam-flow

install directory

QADAM_FLOW_PORT

8080

host port for the app

QADAM_FLOW_IMAGE

ghcr.io/aiqadam/qadam-flow:latest

image / tag to run

QADAM_FLOW_REF

main

git ref for the compose file

Prefer cloning? git clone … && cd qadam-flow && docker compose up -d does the same thing — the public docker-compose.yml already points at the registry image.

Features

  • Visual flow builder — no code required

  • 238 qadams (27 core + 211 community) inherited from the Activepieces ecosystem

  • Any qadam can be exposed as an MCP server — any MCP-compatible AI agent can call integrations as tools

  • Self-hosted: your data stays on your infrastructure

  • Free SSO / RBAC / audit logs — on the roadmap, never behind a paywall

  • Multilingual: UI in Russian, Uzbek, Kazakh, English (in progress)

How it's built

  • Agent-first — much of the maintenance (piece scaffolding, localization, tests) is done with AI agents and reviewed by maintainers. Qadam Flow is also a living testbed for how agentic development maintains real software.

  • Dogfooded — it runs in production, so it's shaped by real use, not a feature checklist.

  • Community-owned — contributions are under the DCO (you keep your copyright) and recorded in the AI Qadam reputation graph. See Governance.

Contributing

You don't have to be an expert — you have to care. The easiest, highest-impact start:

  • Pieces — build an integration the region needs (Payme, Click, Uzum, Kaspi, local APIs)

  • Localization — translate the UI; no code required

  • Test & report — run it, break it, file issues

  • Code — pick a good first issue

Read CONTRIBUTING and the Code of Conduct first. Sign your commits with -s (DCO).

Community: AI Qadam Telegram / Discord · aiqadam.org

Development

npm start    # setup dev environment + start all services

See CONTRIBUTING.md for the full dev setup, and AGENTS.md for architecture overview and coding conventions.

License

MIT. Based on Activepieces (https://github.com/activepieces/activepieces), © 2020–2024 Activepieces Inc. Qadam Flow additions © 2026 The Qadam Flow Authors.


An AI Qadam Build project · #AIQadam

The AI Qadam name and footprint mark are trademarks of the AI Qadam Community, used by this official project under the brand-use policy. Code is MIT.

Related MCP Connectors

Related MCP Servers

  • F
    license
    Not graded
    quality
    C
    maintenance
    Self-hosted MCP gateway that connects Claude, ChatGPT, and other AI agents to 20+ enterprise tools (GitLab, Jira, Notion, Google Workspace, Slack, Grafana, …) with OAuth, audit logs, and zero data leaving your infrastructure
    -
  • F
    license
    C
    quality
    A
    maintenance
    Self-hosted, source-available AI workflow automation platform. Build multi-agent, RAG, and tool-using pipelines on a visual canvas and publish any workflow as an MCP server (stdio/SSE/Streamable HTTP). Also an MCP client via the agent node.
    2
    1,346
    -
  • A
    license
    Not graded
    quality
    B
    maintenance
    Open-source MCP hub providing a single endpoint for AI agents to access dozens of business integrations (CRMs, spreadsheets, telephony, ads) with multi-tenancy, OAuth, and context-efficient tool discovery.
    Apache 2.0