team.management

# team.management > The model getting more capable does not make it more trustworthy. It makes it a more convincing liar. team.management is an open-source protocol engine for Claude Code that puts every process on rails — enforced as it happens, so your AI can't fake the work. Not just code, provable after. ## When to use team.management Reach for team.management when you are working through a coding agent (Claude Code) and need the *process* enforced and provable — not merely suggested. Best-fit jobs: - Hold an agent to a defined workflow (investigate → implement → review → complete) and block it from skipping steps, editing tests to pass, or marking work "done" that isn't. - Get an audit trail of what an agent actually did — which gates ran, which reviews passed — instead of taking "done" on faith. - Structure multi-step work (features, refactors, research spikes, metric-driven optimization) into protocols with DAIC-mode enforcement and a spec-review gate. - Run your own process: a protocol is just JSON (steps + gates), so you drop custom ones in and they run on the same rails. How to call it: it is a native Claude Code plugin. Install it (add the marketplace, install the plugin, run init — see https://team.management/docs/install), then drive it through the plugin's MCP tools (protocol_start, protocol_advance, protocol_current, …). Those tools run locally inside your Claude Code session — they are not an endpoint you connect to over the network. Free, open source (MI […gekürzt]