canopy-sync

Coordinates shared workspace actions through a risk-classified sync protocol for Canopy MCP sessions.

Updated Jun 12, 2026
One-click install
npx skills add https://github.com/UniverLab/univerlab --skill canopy-sync-univerlab
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: canopy-sync
Source: https://github.com/UniverLab/univerlab/tree/main/public/.well-known/agent-skills/canopy-sync
Command: npx skills add https://github.com/UniverLab/univerlab --skill canopy-sync-univerlab

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When working in a shared workspace, uncoordinated file writes, test runs, and commits create conflicts and messy history. This Skill defines an action-driven sync protocol so every operation follows the right coordination steps, whether you work solo or alongside other agents. ## Core Features & Use Cases - Risk-based action protocol: Classifies actions as low, medium, high, or breaking risk and maps each to the correct sequence of sync_get_context, sync_declare_intent, sync_broadcast, and sync_report_status calls. - Tool discovery via get_tools: Returns the right tools and protocol for scenarios like session_start, file_write, test_run, close_session, and multi_agent. - Conflict-aware writes: Before modifying files, checks workspace context and declares intent so concurrent agents avoid collisions. - Use Case: Before refactoring a database layer in a Canopy workspace, declare a coarse mission, execute the edits, then report a stable status so the session history stays clean for the next agent. ## Quick Start Ask the agent to check for conflicts and declare intent before modifying files in the Canopy workspace, then report status when the task is complete.

Frequently Asked Questions about canopy-sync

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I avoid conflicts when multiple agents edit the same workspace?

Classify each action by risk before executing. For file writes, call sync_get_context to check conflicts, then sync_declare_intent, execute the change, and finish with sync_report_status so other agents see your intent in real time.

What is the Canopy action-driven sync protocol?

It is a protocol that maps action risk levels to tool sequences: reads execute directly, tests broadcast before and after, writes follow context-check, declare, execute, report, and breaking changes add impact="breaking" to the intent declaration.

Does canopy-sync work without the Canopy MCP server?

No. The skill requires the Canopy MCP server in the session (requires = "canopy" in skills.toml). Without those sync tools the skill is inert and should not be installed on hosts that never run Canopy.

Should I sync even when working alone in a workspace?

Yes. Sync is based on the action you take, not the number of participants. Solo agents still build clean history that future sessions and other agents can read, so declare, execute, and report regardless.

When should I broadcast messages during a task?

Broadcast milestones only, such as "Completed auth refactor, all tests passing". Avoid dumping every commit message or intermediate step, and keep missions coarse-grained rather than declaring one per file edit.