distributed-mesh

To synchronize agent state across machines, this skill configures a distributed communication mesh that materializes remote squad state locally using git and published contracts. It generates a root mesh.json and sync scripts so agents can read remote filesystem context before operating.

Zone-based distributed coordination separates agent communication into Zone 1 for local state, Zone 2 for remote-trusted state via shared git, and Zone 3 for remote-opaque state via published SUMMARY.md files. This structure allows squads to operate safely across different trust levels.

To set up a shared git repository for distributed squad coordination, provide participating squad names, zones, and connection details to the skill. It will generate a mesh.json configuration, copy bundled sync scripts, and optionally initialize the Zone 2 state repo structure.

Yes, you need jq, git, and curl installed to configure a distributed agent mesh. These dependencies are required to execute the generated sync scripts and manage the cross-machine state synchronization effectively.

Yes, cross-machine agent coordination applies directly to CI, VM-based, and cross-org collaboration scenarios. The distributed mesh allows agents operating in these environments to materialize remote state locally before executing their designated workflows.