How is VibeGov different from AGENTS.md, Cursor rules, Claude skills, or a prompt library?
Agent rules and prompt libraries usually shape one assistant's behavior.
VibeGov gives the repo a broader operating model:
- what work mode the agent is in
- where intent and specs live
- what counts as evidence
- how blockers are reported
- how review and completion claims stay honest
VibeGov can still publish into provider-native files like AGENTS.md or rules directories. The difference is that those files are projections of a shared governance source, not the whole system.
Deeper docs: