Hi HN, I'm working on a context layer built on OSS and open standards that lets agents search your company knowledge and use MCP tools all through a single gateway with code mode. Building this because we believe context can and should be portable across harnesses.

The metric we are solving for here is "cost to resolution". Given any question or a task, how quickly and cheaply can we put an agent on the correct path. Tech behind that:

1. Context map: software to create and maintain a context map of people, teams, systems, relationships, etc. It's built on OKF, import/export from this open standard.

2. Code mode: agents explore the map with a token efficient context eng. method (open source runtime, find portofcontext/pctx on GitHub)

3. Indexing: Hybrid keyword and semantic search on any unstructured text, with code mode on top (we will open source this part soon, although it's nothing special, the interesting part is placing content in the map and putting code mode on it)

Some proof that it works:

1. Goose + code mode: https://aaif.io/blog/how-port-of-context-eliminated-producti...

2. tau-bench: https://portofcontext.com/blog/context-layer-platform-benchm... With Port of Context

To try it, paste this into your coding agent: Set up pctx.ai/onboard.md

It will help you create an account, scan your active MCP connections, and help configure the gateway.

Hope you find this useful and feel free to reach out: patrick@portofcontext.com