docker/docker-agent

This page introduces an external open-source repository. It is not an HDATF product.

What it does

docker-agent is an AI agent runtime from Docker that defines agents in a YAML file and runs them with the docker agent command. The YAML lists the model, instructions, toolsets, sub-agents, fallback models, an iteration limit and skills. Tools are attached as MCP servers through the Docker MCP Gateway, local stdio or remote HTTP, alongside built-in tools such as think, todo and memory. Memory is kept in a SQLite file in the user's home folder, and RAG supports search that mixes BM25 with embeddings, plus reranking. Models come from external providers such as OpenAI, Anthropic, Gemini and Bedrock or from local models in Docker Model Runner, and the user sets the API keys. It installs as a plugin bundled with Docker Desktop 4.63 or later, or through Homebrew and release binaries. By default it asks the user before running tools with side effects, and permission rules can be applied in deny, allow, ask order. The --yolo option approves every tool automatically, sandboxed runs need Docker Sandboxes installed separately, and the working folder is mounted with write access. It also hands work to CLIs on the PATH such as claude and codex, serves a REST API with a session database, shares agents through OCI registries and evaluates by replaying saved sessions. Telemetry is on by default, and the documentation says not to assume the collected data is free of secrets or personal information.

License

Apache-2.0 Permissive, with a patent grant. Commercial use is allowed; keep the notices and state your changes.

More in this category

• omnirush-ai/omnirush-gui
  OmniRush is a desktop application for asking a coding agent to work on a selected project. Prompts and project files become inputs, while edited files, command results and Git work become outputs. An Electron shell starts a local server and a bundled OpenCode engine that exposes skills and MCP tools to the selected model. MCP is a common interface through which agents call external tools. The application requires an approved OmniRush account and device sign-in before it runs prompts, and the available model catalog comes from that account. Published desktop builds target Apple Silicon macOS and Linux, not a native Windows desktop release. Free access is offered by the service itself, but its privacy policy permits collecting session traces and project uploads for model development and research partners. The desktop code is largely MIT-licensed, while the enterprise directory has separate terms, so neither local execution nor public source means private processing or unrestricted enterprise reuse.
• morluto/rea
  Gives agents tools to investigate application behavior and native binaries. Records evidence and distinguishes static inference from runtime observation.
• CopilotKit/OpenDots
  OpenDots is a self-hostable template for persistent AI agents with individual computers, spanning text, calls and Slack. Its early-stage workspace includes working-document Spaces and in-chat computer views and human review cards.
• anteloc/ldraw-nova
  ldraw-nova lets an AI agent build LEGO models from an idea and user guidance. Outputs include LDraw source, multiple views, editable glTF files and chat history.
• devswha/herdr-web-ui
  A browser and phone client for herdr that opens the agent sessions already running on a machine and shows each agent pane as both a chat and a live terminal. It attaches screenshots, mentions files and queues messages while an agent works, opens a path from the chat in a file viewer, browses a pane's folder, adds Linux and macOS machines over SSH, and sends web push alerts.

Only repositories in the ranked Trendshift lists are included, and the lists are used only to find candidates. We do not copy their ranks. Descriptions, licenses and star counts come from each GitHub repository. The notes are our own reading. We have not tested these projects, and a place on a trending list does not prove quality.