Back to Ecosystem Pulse
ToolProduction ReadyMCP
agor
by preset-io
Multiplayer canvas for orchestrating and auditing coding agents
TypeScript
Updated Jul 25, 2026
Share:
Overview
Orchestrates multi-model coding sessions (Claude Code, Codex, Gemini) on a shared, multiplayer canvas. This follows the Tool Use Pattern.
Key Benefits
As agents collaborate on code, visibility into which agent made what suggestion and how it affected the repo becomes essential for trust and debugging. Agor captures agent interaction logs and maps them to git worktrees so you can audit decisions, detect failure modes, and build an agent track record. That contextual linkage is valuable for eventual agent-to-agent evaluation and continuous reputation signals. This contextual linkage aligns with the Mutual Verification Pattern.
Target Use Cases
Teams or developer tools projects that need collaborative, auditable multi-agent coding sessions tied to git worktrees. This approach supports planning and orchestration, aligning with the Planning Pattern. It also benefits from discovery and capability matching in complex tool ecosystems via the Capability Discovery Pattern.
Real-World Examples
- Audit agent suggestions by mapping AI conversations to git worktrees and file diffs
- Run collaborative pair-programming sessions with multiple LLMs and visualize agent actions in real time
- Investigate agent failure modes by replaying interaction history alongside code changes
Works With
anthropicopenaigooglegit
Topics
agent-orchestrationagenticai-agentsai-codingai-pair-programmingcanvasclaude-codecodexcoding-assistantdeveloper-tools+9 more
Similar Tools
autogencrewai
Keywords
multi-agent orchestrationagent track recordagent interaction loggingmcp