Proposal: inter-agent collaboration & Cotal evaluation#623
Closed
bborn wants to merge 1 commit into
Closed
Conversation
Evaluates whether to adopt Cotal (github.com/Cotal-AI/Cotal) for live inter-agent collaboration. Recommendation: adopt Cotal's *model* (multicast/ unicast/anycast + presence + A2A AgentCards) on taskyou's existing substrate (SQLite + MCP server + events bus) rather than taking a hard dependency on its TypeScript/NATS runtime, which contradicts the single-binary, CGO-free, local-first value prop. Grounds the design in the lessons from the closed PR #357 (relay): replaces its fragile idle-detection delivery with pull-based MCP tools, derives presence from task status (no manual register/unregister memory leak), and addresses by task id/channel/role instead of task-title-derived names. Keeps a Cotal bridge as an optional, config-gated Phase 3 adapter for interop. No runtime behavior change — adds the proposal doc only; seeking a direction decision before implementation. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Answers the question: can we leverage Cotal in a smart way to enable better inter-agent collaboration?
This PR adds a proposal document only — no runtime behavior changes. It's seeking a direction decision before any implementation.
Recommendation in one line
Adopt Cotal's model, not its runtime. Build native MCP-first agent messaging + presence on the infra taskyou already has (SQLite + the MCP server + the events bus), keep a Cotal bridge as an optional config-gated adapter, and avoid a hard dependency on Cotal's TypeScript +
nats-serverstack — which would contradict the single-binary, CGO-free, local-first value prop.What's in the doc
AgentCards; and an honest table of what its Node/NATS runtime costs us.relay) — keeps its messaging idea but replaces the two things that made it inorganic/unmergeable: fragile idle-detection injection (→ pull-based MCPtaskyou_inbox) and a manual agent registry that leaked (→ presence derived from task status). Addresses by task id/channel/role instead of task-title-derived names.agent_messagestable, four MCP tools (taskyou_send_message,taskyou_inbox,taskyou_roster,taskyou_claim), and how messaging makes the existing dependency primitive feel organic by driving auto-unblock.📄
docs/superpowers/specs/2026-06-25-inter-agent-collaboration-cotal.mdDecision needed
See §8 (Open questions). The key one: approve building Phase 1 (native unicast messaging + roster via MCP) as the next step, keeping Cotal as an optional Phase 3 bridge?
🤖 Generated with Claude Code