Repository navigation
id(phase-14): Agents in worlds (MCP bridge) — core topic (research + design) #6280
Description
Agents in worlds (MCP bridge) — Phase 14 core topic (research + design)
Filed from the next-horizons discussion #1240, decided by Drew on 2026-11-05: all themes are planned. This issue is the core topic for Agents in worlds (MCP bridge). Its job is research and design — broaden the idea beyond what is already understood, set up multiple next steps, and produce implementation plans. No implementation starts here; the output is design documents and implementation issues.
Drew's principles for everything in Phase 14 (from #1240, 2026-11-05)
- Local-first is always the goal, AND a hosted version is intended: central coordination supporting a large graph of broadly accessible objects.
- Anything built must be completely usable by others hosting a small or large server themselves. P2P-hosted worlds and data behave the same as a large hosted public server; differences live in backend scale elements and access controls.
- Users connect to each other and transfer/interact directly wherever possible, so hosted resources are an offloadable backstop, not a requirement. The app must not technically need us to function. Collaborative / shared / federated / local-first / p2p.
- Our hosted instance is essentially public; other operators open theirs only as much as they want. Publishing paths: share to a big server, publish on pkarr, or self-host. All comms through iroh on the backend; first-class clients.
The idea, broadened
The most differentiating theme: a world is a place, and agents are participants. Broaden beyond an MCP wrapper: agent identity and roles inside a world, what an agent may see (screens are the natural interface — an agent shell is just another conformance-checked client), capability ceilings as the safety model, audit of every agent action (F12.6), and the economics (F12.10 quotas) of an always-on agent. Prototype cheap, design deep: the spike proves the loop; the design decides what agents are for in a world (helpers, players, archivists, bridges to other systems).
Research questions (first pass)
- Agent as shell: reuse the F13.1 contract — an agent client that reads screens and emits input; what breaks (latency, attention, partial views)
- Identity & roles: agent keys, world roles for agents, owner accountability, revocation
- Guardrails: capability ceilings per agent, fuel/quotas per action, audit trail shape (F12.6 events an owner can actually review)
- MCP surface: which world operations become tools; what must NOT be a tool
Dependencies and existing pieces
Builds on Phase 13 shell contract, F12.3 authority (#835), F12.6 audit (#1063), F12.10 quotas (#1067). Agreed Phase 14 spike: prototype alongside hosted-worlds/discovery design.
Next steps — the version ladder
- v1 — first-iteration design + spike plan — its own issue, filed in the same pass and linked in the first comment below
- v2 — outline: agents as citizens — its own issue, filed in the same pass and linked in the first comment below
- v3 — outline: shared agent worlds — its own issue, filed in the same pass and linked in the first comment below
- v4 — outline: hosted agent services — its own issue, filed in the same pass and linked in the first comment below
Plus a standing rule for this topic: anything the research surfaces that doesn't fit v1–v4 gets its own issue (Drew's rule: every feature, concept, or idea gets an issue), and future investigations are listed as comments here rather than lost.
Sequencing
Design queue order: F12.4 provider registry (#1061) and F12.10 scheduling/quotas (#1067) designs land first; Phase 14 v1 designs then run hosted worlds → discovery → agents spike → the rest. Implementation follows Phase 13. The drop rule applies throughout: anything that proves blocked, wrong, or superseded gets parked or closed with a note — never a silent stall.
Acceptance criteria
- A research + design output exists for Agents in worlds (MCP bridge): design documents broadening the topic beyond the current understanding, per the Phase 14 principles in the body (local-first plus hosted coordination).
- Multiple concrete next steps are produced as implementation plans / implementation issues filed from this topic.
- No implementation starts under this issue itself (per its own charter).
Activity
developing-today commentedon Nov 5, 2026
CODE-1846
dezren39 commentedon Nov 5, 2026
Version ladder for this topic (filed in the same pass, 2026-11-05):
Phase 14 plan of record: #6312. Discussion of record: #1240.
dezren39 commentedon Nov 6, 2026
Intake review
🤖 Fleet monitor intake (repo-fleet-monitor), 2026-11-06.
dezren39 commentedon Nov 6, 2026
🤖 Agent note · handed off
dezren39 commentedon Nov 6, 2026
agent-note:v1
task 11052 picked up by lane c29174f1 (openchamber scheduled lane, worker repo-worker-code): triage of this Phase 14 core topic (research + design) at current main. Checking ladder, ordering blockers, and deliverable state.
dezren39 commentedon Nov 6, 2026
Triage findings (verified fresh 2026-11-06 ~14:58Z via gh + main @ 2bcd46e)
Live and correctly scoped — not actionable yet. Research/design core topic under the Phase 14 plan of record (#6312, OPEN). Charter forbids implementation here; outputs are design docs + implementation issues.
.opencode/plans/ordoc/at current main. Nothing fixed, superseded, or stale — the drop rule does not apply.Disposition: triaged. This issue stays open as the topic tracker; it should not be re-queued until #1061/#1067 designs have landed and the hosted-worlds/discovery designs it spikes alongside are moving.