Choose theme
AI Resources
OpenAI MCP Extensions
OpenAI MCP Extensions provides SDKs and a specification for adding ChatGPT-specific capabilities to MCP plugins. Builders can connect an app to sidebar entrypoints, custom file viewers, searchable mentions, and native settings.
The repository includes TypeScript and Python SDKs plus a Bits & Bolts example plugin. The features available to an app depend on its host. Use this as a first read, not a recommendation. Open the original project before trusting details like terms, limits, privacy, cost, setup, or safety.
What it is
SDKs for richer MCP plugins
Add interfaces and resource interactions around the tools your MCP server exposes. The specification documents the messages and metadata that connect those pieces.
Why it stands out
Several ways to reach an app
A plugin can register a sidebar entry, handle selected file extensions, or offer resources through composer mentions. Each entry gives the user a different way to bring work into the app.
Availability
Source, packages, and an example
The TypeScript package is on npm and the Python package is on PyPI. OpenAI declares Apache-2.0 for the repository; its SDK releases are at an early stage.
Why it matters
What makes it useful
For a parts library, a builder can let someone browse objects visually and choose a reference for a message. The Bits & Bolts example shows this with CAD parts and a 3D viewer, so a user can point to the object they mean.
What to know
Where it fits
The TypeScript SDK extends the MCP server and MCP Apps SDKs, with separate server and app imports. Python provides server-side extensions; the app interface still uses the TypeScript package. Your plugin supplies its own tools and interface.
Notable points
What stands out
File handlers can also become editors. The specification restricts resource writes to the opened file when the host marks it writable. A supplied ifMatch version token makes a stale edit return a conflict instead of saving over a newer version.
Before using
What to review
The TypeScript guide says extension fields may remain unavailable when the current host does not support them. Code needs to handle that case.
Building the Bits & Bolts example from source requires Node.js 22 or later and pnpm. Its packaged output still requires Node.js.
Settings storage belongs to the MCP server. Adding the settings interface does not provide persistence for your application.
Reader fit
Who may find it relevant
MCP server developers and teams building plugin interfaces can use the SDK examples as implementation references. Readers curious about plugin interaction can explore the demonstration; adapting the code requires development experience.
Editorial note
Why LifeHubber lists it
A plugin interface can keep a person's selection and the assistant's context connected: the model-context API lets the app share selected items as structured data. This earns a place here for builders who need the assistant to discuss the object a user has chosen, with an explicit data handoff they can implement.
Source links
Source materials
Reader note
Before relying on this entry
LifeHubber lists entries to help readers inspect AI projects, not to endorse them or prove they are safe, suitable, accurate, maintained, or right for a specific use. We do not verify every entry in depth. Before relying on anything listed, review the original materials, terms, privacy practices, limits, and risks that matter for your situation.
More in AI Agents
Keep browsing this category
Explore more AI agent projects.
Paperclip
paperclipai/paperclip
A self-hosted server and dashboard for coordinating agent teams through companies, goals, roles, issues, heartbeats, budgets, approvals, and persistent activity records.
OpenDots
CopilotKit/OpenDots
An early CopilotKit template for a persistent specialist-agent workspace with editable pages, per-Dot computers, and separately configured conversation, voice, and Slack services.
MaxKB
1Panel-dev/MaxKB
An open-source knowledge-base agent platform for building RAG-backed Q&A assistants and agent workflows, with document upload and web-source collection, vector/full-text/hybrid retrieval, model-provider connections, MCP/tool use, web embedding, Docker/offline deployment paths, and GPL-3.0 licensing.
For project maintainers
Listed here? You can use the badge.
If you maintain a project with a current LifeHubber listing, you may add the optional “Listed on LifeHubber AI Resources” badge to its README, docs, or website. No introduction or permission request is needed.
