trovex vs Continue.dev
Continue is a full assistant with its own retrieval. trovex is a context source you plug in. They compose — here's the honest comparison.
Continue.dev is a free, open-source AI coding assistant and framework — chat, autocomplete, custom models — with its own codebase retrieval and MCP support. trovex is an MCP context server that serves the one canonical doc per query with a freshness marker, for about 60% fewer tokens per lookup. They're at different layers: because Continue can connect MCP servers, you use trovex inside Continue rather than instead of it. trovex sharpens the doc-context part; Continue stays your assistant.
What is Continue.dev?
Continue is an open-source AI coding assistant for VS Code and JetBrains, plus a framework for building custom retrieval pipelines. It brings chat, inline edits, and autocomplete with the model of your choice, and it has its own @codebase context and MCP integration. You pay only for the model you use.
Where trovex fits
trovex isn't an assistant — it's the thing that answers "which doc is canonical?" cheaply and correctly. Continue's built-in retrieval returns relevant chunks; trovex returns the single current doc, marks stale and duplicate copies so they're skipped, reads at the section level, and lets agents write records back. Point Continue at trovex over MCP and your repo's canonical markdown becomes part of its context.
| Capability | Continue.dev | trovex |
|---|---|---|
| Type | AI assistant + RAG framework | MCP context server |
| Chat / autocomplete / edits | ✓ | — not its job |
| Doc retrieval model | Relevant chunks (@codebase) | ✓ one canonical doc |
| Freshness signal | — | ✓ canonical / stale / duplicate |
| Write-back (SSOT) | — | ✓ shared write path |
| Works via MCP | ✓ MCP host | ✓ MCP server (plug in) |
| License / cost | OSS, pay for the model | OSS (AGPL core / MIT CLIs), local |
When is Continue the right choice?
When you want a full, customizable assistant in your editor — chat, autocomplete, your own models — Continue is an excellent OSS choice. Add trovex when you want your repo's docs served as one canonical, current answer rather than chunks, with a freshness signal and a shared write path. The two layer cleanly; this isn't an either/or.
FAQ
Is trovex an alternative to Continue.dev?
Not exactly — different layers. Continue is a full OSS assistant and framework with its own retrieval; trovex is an MCP context server serving one canonical doc per query. Continue can connect MCP servers, so you use trovex inside Continue rather than replacing it.
How is trovex's context different from Continue's @codebase retrieval?
Continue returns relevant chunks; trovex returns the single canonical doc with a freshness marker so stale/duplicate copies are skipped, and lets agents write records back. It's opinionated for a repo's markdown rather than general code RAG.
Can I use trovex and Continue together?
Yes. Keep Continue as your assistant and connect trovex as an MCP server so your repo's canonical docs are available to it. They're complementary.
Sharper doc context for your assistant.
trovex is in private beta — request access and plug it into Continue or any MCP client.
Open source (AGPL-3.0 core, MIT CLIs). Local-first — your docs never leave your machine.