Updated
trovex for small teams
On a small team, every dev's coding agent re-derives the same things and your docs drift out of sync. trovex gives the whole team's agents one shared point to read from and write to — a single source of truth, no copies to sync. Local-first, no hosted account.
The problem at team scale
Two developers, two agents, one repo — and three versions of "how deploys work." Each agent reads whichever doc looks most relevant, re-derives what the other already figured out, and saves notes into its own context that no one else sees. Multiply by every runbook, ADR, and README that overlaps, and the team is quietly paying for the same lookups over and over while working from copies that disagree.
One point of passage for the whole team
trovex closes that loop. It indexes the team's docs and serves the one current doc per query with a freshness marker, and every agent writes what it learns back through a single shared point — so a fix one dev's agent finds is readable by everyone else's next session. No copies to drift, nothing to sync. That's what keeping multiple agents on the same docs actually requires.
How it rolls out
Bottom-up, no procurement. It's a local-first MCP server each developer runs against the team's repo — no SaaS to buy, no SSO project, no data leaving your machines. One person installs it, sees the savings, and it spreads.
When scaling it well is the hard part
Adopting the tool is a one-minute install. Making a whole fleet of agents reliable and consistent across a team — scoped permissions, review gates, shared context that stays current — is a harder problem. trovex is open source you can run yourself; if you want hands-on help getting there, the founder offers consulting, and trovex is the proof of how it's done. Start with the tool; the help is there if you need it.
FAQ
How do I keep a team's coding agents from drifting out of sync?
Give every agent one shared point to read and write, instead of each maintaining its own copy. When one agent records what it learns through a single store, every other agent and teammate reads that record back. With one point of passage there are no copies to drift and nothing to sync — a single source of truth by construction.
Does trovex need a server or cloud account for a team?
No hosted account. trovex is a local-first MCP server each developer runs against the team's repo. No SaaS to buy and no procurement — adoption is bottom-up, one developer at a time, and the docs never leave your machines.
We're rolling agents out across the team and it's messy. Can someone help?
Yes. trovex is open source you can adopt yourself; if you want hands-on help making a fleet of agents reliable and consistent across the team, the founder offers consulting, and trovex is the proof of how that's done.
Give your team's agents one source of truth.
One shared read-and-write point. No copies to sync. About a minute to set up.
uv tool install git+https://github.com/TsukumoHQ/trovex
Open source. No cloud, no API keys. Your docs never leave your machine.