Your questions,
clearly answered.

ConductLoop is a free, open-source AI workspace for macOS. It brings project agents, tasks, models and local context together through Brain, Boardroom and MCP. Hierarchical multi-agent integration is in preparation.

The current app includes Brain chat with cited sources, Boardroom project agents and tasks, an MCP server, local search, Obsidian sync, capture and briefings. The planned CEO-led team extends this foundation.

Yes. The desktop app — the Python engine, browser UI and macOS shell — is public on GitHub under the MIT license. You can read it, build it yourself, fork it and redistribute it under that license.

The signed desktop app is macOS (Apple Silicon) for now. The engine is plain Python and SQLite with a browser UI at http://127.0.0.1:7777, so running it from source on Linux or Windows should work — but it isn't tested there yet. Reports and pull requests are welcome.

No. The public download does not require registration, email or a ConductLoop license key. An external model provider may require its own account.

The published download targets macOS 13 or later on Apple Silicon. This website does not offer a native installer for Intel Macs, Windows or Linux.

Begin with a small project and sources you want to index. Ask a specific question in Brain and review its citations. Add sources when you know which information is missing.

The public EVMCOD/conductloop-releases GitHub repository contains releases. The download page links to the latest release and the home page includes the release feed.

You will build your team and direct the CEO agent. The CEO will orchestrate the work of the other agents around your goal. This hierarchical integration is in preparation and is not included in the current v1.2.0 download; no release date is announced here.

In the planned model, you set the goal and direct the CEO. The CEO coordinates team agents. Autonomy and permission details have not been announced; CEO describes an AI coordination role.

No. Hierarchical integration is in preparation and is not included in the current v1.2.0 download. The website illustration is conceptual.

No release date is announced on this website. Available capabilities will be documented in release notes when published.

A role catalogue and configuration have not been announced. The blog proposes research, writing and review examples to help define responsibilities; these are not bundled app templates.

The app works with supported providers, but per-agent model configuration within the future hierarchy is not confirmed here.

Boardroom organizes project agents, tasks and runs. The planned integration adds the hierarchy described on the website: you direct the CEO and the CEO orchestrates the team. Current features and that planned development are distinct.

Coordination should not be treated as a guarantee of accuracy. Review important results and check their sources. The planned hierarchy keeps a person in charge of the goal.

It's free: no trial, subscription, account or email gate. You pay your model provider directly if you use a cloud model (Anthropic, OpenAI or MiniMax), or nothing for model usage with local Ollama. ConductLoop itself does not proxy model calls.

Anthropic Claude, OpenAI GPT, MiniMax and any local model served by Ollama. Bring your own key — no markup, no proxy. You can switch provider at any time; the memory stays the same, because the memory is the asset and the models are replaceable.

You configure your model-provider key so ConductLoop can use that service. Usage is billed under the provider’s terms; ConductLoop does not proxy those model calls.

No. The app is free, but a cloud provider may charge for usage. Consult its pricing and usage dashboard. Local Ollama uses your machine’s resources.

Local storage and search do not require cloud-hosted files. Cloud-model calls, watched URLs and external integrations require network access. With an already installed local Ollama model, model calls are processed on your machine.

Speed depends on the model, hardware, context size and machine load. Narrow a query and consult Ollama documentation to review resources and configuration. There is no guaranteed speed across all machines.

Yes. ConductLoop ships an MCP server (ops mcp) with search, recall and capture tools, so any MCP client can read from and write to the same memory. It also syncs both ways with your Obsidian vault as plain markdown.

No. ConductLoop can index notes, repositories and other project sources. Obsidian is an integration for working with a vault and syncing memory through Markdown.

The local index lets you query selected sources. Original notes and repositories remain in their locations; Obsidian sync uses Markdown files. Keep your usual backups.

ConductLoop’s MCP server exposes search, recall and capture tools to compatible clients, making context available to other tools. MCP access alone does not create the planned agent hierarchy.

Check that the relevant source is included in your indexed material and ask a more specific question. Review citations before accepting an answer. If the material does not contain the fact, supply a source or treat the answer as uncertain.

No. A citation helps trace a source, but you should check that it supports the claim and is current. Documents can also contradict one another.

The index and app data stay in local SQLite. Prompts and selected context go to the model provider you configure; with local Ollama, model calls stay on your Mac. Configured integrations may contact external services. Error reporting only runs if configured.

Open an issue for bugs and feature ideas, or send a pull request — CONTRIBUTING.md in the repo explains how to set up the dev environment and run the tests. Starring the repo also helps more people find it.

Not necessarily. The index is local, but prompts and selected context are sent to your configured cloud model. External integrations may also make network calls. With local Ollama, model calls stay on your machine.

Describe the issue, app version, system and reproduction steps. Before attaching logs or screenshots, remove keys, tokens and unnecessary private content. The public app repository accepts issues.

The app is published under the MIT license. Consult the repository’s LICENSE file for its conditions and notices that must be preserved.

Read the blog guides, open an issue in the public app repository or email hello@conductloop.com. For reproducible bugs, include minimal technical context and avoid sharing private data.