Software Development
Give your AI tools the review standards, architecture context, and writing rules your team already relies on, so every session starts productive.
OpenAI Codex MCP servers
Codex CLI, the IDE extension, and the desktop app share one MCP config. Add https://knotr.ai/mcp/v1 as a Streamable HTTP server, then authenticate. This stays separate from ChatGPT. Briefs worth keeping can live on as a link you can open later.
Setup takes a minute after you create a profile. Open Integrations → MCP.
Knotr in OpenAI Codex, not a generic feature list.
Load your Knotr profile once. Codex CLI, the IDE extension, and the desktop app read the same MCP config.
The workflows you already keep in Knotr are ready in Codex after you connect. You do not rebuild them in a second tool.
Ask Codex to save a brief or page as a Knotr artifact. Group several into a Collection when you want one link someone can walk through in order.
Exemplary steps only. After you sign up, Integrations has the real connect buttons for this profile.
OpenAI Codex shares MCP setup across Codex CLI, the IDE extension, and the desktop app. Add a Streamable HTTP server at https://knotr.ai/mcp/v1 in ~/.codex/config.toml. In the desktop app, open Settings → MCP servers, add the server, then choose Authenticate. In the IDE extension, open the gear menu, choose MCP servers, add the server, then choose Authenticate. This is separate from the ChatGPT Integrations tab.
After you sign in, Codex can use your profile, the skills you share, and the notes you attached. It can also save a brief or page you want to keep.
Pair OpenAI Codex with a workspace tuned to how you work.
Give your AI tools the review standards, architecture context, and writing rules your team already relies on, so every session starts productive.
Keep strategy, metrics, and prioritization rules in one place so PRDs and stakeholder updates stay aligned with how your team actually works.
Carry market assumptions, strategic themes, and decision frameworks into every planning session.
Same knowledge layer, every supported client .
Your voice and skills in ChatGPT, plus a place to keep the work that would otherwise die in the chat.
Claude Desktop, Claude (web & mobile), and the Claude API, from the Knotr plugin or the directory connector.
The same profile and skills as Cursor, from the CLI, with a place to keep what the session produces.
Your standards in every repo, and canvases you can reopen and share after the chat is gone.
Your profile in Gemini Spark through a custom Connected App. Gemini Chat cannot use that connection.
Your profile in Google Antigravity. Add the custom remote server and sign in. Separate from Gemini Spark.
Your profile and skills in Grok. Add the remote MCP server, sign in, and keep the work worth saving.
Any compatible client. Same profile, same skills, same place to keep the output.
Your persona in the Muse app. Skills and saved work live in Muse Code.
The same profile and skills as Claude Code, from Muse Code, with a place to keep what the session produces.
Your profile and skills in OpenCode. Add the remote server, sign in, and keep the work worth saving.
Bring your profile into Perplexity and keep the work worth saving. Add Knotr as a custom connector and sign in.
Sign up free, create a profile, then open Integrations → MCP.
Free · no credit card · about 2 minutes