What it does
Treats coding agents as first-class workspace participants rather than as APIs or tools. Each agent has a profile, appears on the board, posts comments, creates issues, changes status, and reports blockers — interleaved with human team activity in a single timeline.
Mechanics from the captured pages
- Assignee dropdown contains both humans and agents. Quote: "Pick your agent from the assignee dropdown — just like assigning to a teammate."
- Autonomous participation: agents create issues, leave comments, and update status on their own — "not just when prompted."
- Unified activity timeline: "One feed for the whole team. Human and agent actions are interleaved, so you always know what happened and who did it."
Example transcript from the Multica landing demo (MUL-18 — Refactor API error handling middleware):
Alex Rivera assigned to Claude 3:02 PM
Claude changed status: Todo → In Progress 3:02 PM
Alex Rivera "The current error responses are inconsistent…" (comment)
Claude "I've standardized error responses across 14 handlers. Each
error now includes a code, message, and request_id. PR #43
is ready for review."
Alex Rivera "Looking good. Make sure to preserve existing HTTP status…"
Target ICP
Solo developers + small teams running 1+ coding agents. The framing serves "agent collaborator," not "agent automation."
UX entry points
- Assignee picker on any issue card.
- Activity feed (board-level + issue-level).
- Agent profile page (per agent).
multica issue create/multica issue listfrom CLI.
how the agent's reply latency / streaming is surfaced on the issue card. The screenshot suggests it's a comment-with-timestamp pattern, but inflight-stream UX wasn't captured.
Differentiators (claimed)
- Same UX as human-to-human assignment — no separate "AI tools" interface.
- Vendor-neutral — Claude, GPT, Gemini, Cursor, etc. all assignable through the same dropdown.
- Compound activity history — every assignment is a board entry, searchable and reportable later.
Versus the rest of the vault
| Mechanic | Multica | Replit | Anything | Fly | Hercules |
|---|---|---|---|---|---|
| Agent has a board profile | ✅ | ❌ | ❌ | ❌ | ❌ |
| Agent assigns / comments autonomously | ✅ | 🟡 (Task Board) | 🟡 (Max) | ❌ | ❌ |
| Multi-agent assignee dropdown | ✅ vendor-neutral | ❌ (own agent) | ❌ (own agent) | ❌ | ❌ |
| Issue tracker integration | ✅ first-class | ❌ | ❌ | ❌ | ❌ |
The closest feature in the rest of the vault is Replit's Replit - Task Board, but that's a kanban for the Replit Agent's tasks. Multica generalizes the pattern across any user-installed coding agent.
Sources
- multica.ai homepage (the demo screenshot + transcript)
- github.com/multica-ai/multica (README)