Skip to main content

OpenClaw Mission Control Project Plan

Documentation Map​

OpenClaw Mission Control Project Plan

Pattern Upgrade Program​

The repo-owned Event Modeling, Plane/Gitea materialization, business-capability projection, and autonomous-enterprise planning posture now live in:

  • PLAN_PATTERN_UPGRADE.md

Use that plan for:

  • Pattern/Fabric ownership split
  • Event Modeling wave execution (W0 through WF)
  • the direct post-WF continuation (WG through WJ-HOST)
  • the current governed continuation refs:
    • jhf-pattern#181
    • jhf-pattern#182
    • jhf-pattern#183
    • jhf-pattern#184
    • jhf-openclaw-env#238
  • Pattern materialization invariants and contracts
  • business-capability and autonomous-enterprise extension rules

This document remains the broader Mission Control product/delivery plan.

Purpose​

This document is the master delivery plan for turning Mission Control into a full OpenClaw project-management suite.

Target outcome:

  • Mission Control is the operator-facing control plane for software delivery
  • Plane is a shared human-planning layer, not just a one-way issue mirror
  • Git providers contribute real delivery state: reviews, branches, releases, pipelines, deployments
  • n8n provides project-management automation, not just technical webhook plumbing
  • MCP gives agents a safe, audited interface for PM, Git, and CI/CD actions

Current Program Status​

Current date: 2026-03-28

Overall program status:

  • Foundation and platform hardening: done
  • UX stabilization and browser regression safety: done
  • Phase 1 PM-object foundation: done
  • Phase 2 shared planning and delivery foundation: done
  • Phase 2 live-host validation and hardening: done with one known external limitation
  • Full-suite completion: not done

Known external limitation:

  • The current self-hosted Plane instance does not expose project pages over the API in the same way as cycles/modules.
  • Mission Control now handles this cleanly in the UI:
    • page sync is marked unavailable
    • users are not pushed into a broken button path

What Already Exists​

Completed Base​

  • Stable Inbox, Today, Work, Risks, Recovery, Settings operator surfaces
  • Settings assistant, repo provider flows, Plane connection flows, n8n automation flows
  • Live Gitea, Plane, n8n, and runtime-host verification on <internal-runtime-redacted>
  • Broad backend regression coverage
  • Broad Playwright mock-smoke coverage
  • Broad Playwright live-host coverage

Completed PM/Data Foundation​

  • Planning objects:
    • cycles
    • modules
    • pages
  • Delivery objects:
    • pipeline runs
    • deployments
    • releases
    • environments
  • Delivery provider sync endpoints
  • Shared-edit Plane flows for:
    • cycles
    • modules
    • pages, with graceful capability fallback
  • MCP v1 for PM, delivery, and Git/provider actions

Completed Live Verification​

Live browser coverage now proves:

  • project creation
  • setup-assistant journeys
  • repo/Plane/automation settings flows
  • work and risk flows
  • planning create/delete for cycle/module/page
  • planning sync -> pull -> cleanup for cycle/module
  • page capability fallback when Plane pages API is unavailable
  • delivery create/delete for pipeline/deployment
  • delivery provider sync with visible gitlab cards
  • recovery sync/webhook/n8n dry-run flows

Phase Status​

Phase 0: Core Platform Hardening​

Status: done

Scope:

  • stable backend contracts
  • auth and routing cleanup
  • regression tests for setup/status logic
  • host-aware configuration cleanup
  • browser-safe error surfacing

Delivered:

  • stable settings summary contract
  • regression suite for setup/status rules
  • safer host-path handling
  • cleaned API mismatches and route gaps
  • stable live deployment path

Exit criteria:

  • core API and settings flows no longer silently break
  • browser and backend suites green

Phase 1: PM Object Foundation​

Status: done

Scope:

  • add first-class planning and delivery domain objects
  • add UI surfaces for Planning and Delivery
  • add MCP PM foundation

Delivered:

  • backend models and APIs for cycles/modules/pages
  • backend models and APIs for pipelines/deployments/releases/environments
  • PlanningView
  • DeliveryView
  • MCP tools for planning/delivery primitives

Exit criteria:

  • Mission Control can store and operate on PM objects independently of Plane/Git

Phase 2: Shared Plane Edit + Delivery Ingestion​

Status: done

Scope:

  • shared-edit Plane bridge
  • provider-backed delivery ingestion
  • planning-aware work surface
  • MCP PM/Git/CI-CD expansion

Delivered:

  • cycles/modules shared-edit with Plane
  • pages with graceful Plane capability detection
  • provider-backed delivery sync for pipelines/deployments/releases
  • planning-aware work metadata
  • MCP expansion to PM, Git, and delivery actions

Exit criteria:

  • Mission Control is no longer just a board plus settings; it is a working PM layer

Phase 3: Workflow Completion​

Status: in progress

Goal:

  • turn the implemented surfaces into a complete daily operating system

Still to do:

  • native Plane page-object support later if the target Plane API exposes project pages consistently; the managed Plane-issue mirror already covers the supported shared-edit roundtrip today
  • richer delivery-to-planning links:
    • attach release to cycle
    • attach release to module
    • environment-aware release readiness
  • first-class release blocking/unblocking flow in UI
  • richer MCP write operations for reviews, branches, releases, and escalation
  • better agent/runtime bridge from Mission Control work into actual OpenClaw agent sessions

Exit criteria:

  • every major PM object has a complete operator flow
  • no important action depends on hidden or manual host-side work

Phase 4: Git-Native Software Delivery Suite​

Status: not started

Goal:

  • make Mission Control a serious PM suite for software delivery, not just a sync shell

Still to build:

  • provider-native release readiness dashboard
  • review queues by repo/module/cycle
  • branch hygiene and stale-branch handling
  • PR/MR approval-state ingestion and actioning
  • milestones/labels support where provider APIs offer it
  • deployment and environment health escalation into Recovery

Exit criteria:

  • software-delivery teams can manage real delivery from Mission Control without losing provider context

Phase 5: Automation as PM Orchestration​

Status: not started

Goal:

  • make n8n a planning and delivery orchestration layer

Still to build:

  • cycle kickoff workflows
  • cycle closeout workflows
  • deployment failure escalation workflows
  • release readiness workflows
  • weekly delivery digest workflows
  • project pages/doc freshness workflows
  • workflow attachment to:
    • project
    • cycle
    • module
    • environment
    • release

Exit criteria:

  • project-management automation is configured in product terms, not raw integration terms

Phase 6: Agentic PM via MCP​

Status: foundation done, full capability not done

Goal:

  • allow OpenClaw agents to work against PM state through Mission Control safely

Delivered already:

  • MCP server exists
  • PM/delivery/Git/provider tools are exposed through Mission Control APIs

Still to build:

  • stronger MCP tool ergonomics for agent workflows
  • richer review/release/deployment action tools
  • clearer per-tool audit stories and user-facing traceability
  • skill / MCP guidance for agents to use Mission Control as the default PM interface
  • stronger runtime bridge from PM tasks to agent sessions/mailboxes

Exit criteria:

  • an agent can receive, update, hand off, escalate, and close work entirely via Mission Control/MCP

Phase 7: Production Readiness​

Status: not started

Goal:

  • make the product operationally complete for sustained use

Still to build:

  • backup/restore runbooks for all critical state
  • deployment promotion flow
  • upgrade playbooks
  • environment-specific config strategy
  • telemetry and operator diagnostics for sync failures
  • rate-limits, quotas, and audit retention strategy
  • stronger chunk-splitting/performance work on the frontend

Exit criteria:

  • the system is safe to run as a daily shared PM control plane

Sprint Plan​

Sprint 1: Shared Planning Completion​

Status: next

Scope:

  • complete page strategy
  • finish explicit conflict-state UX
  • add module and page pull/sync visibility improvements
  • add more explicit per-object Plane capability badges

Deliverables:

  • final page strategy decision:
    • full Plane page shared-edit if API allows
    • or Mission-Control-authoritative pages with optional Plane mirror disabled by capability
  • conflict panel with clearer next actions

Sprint 2: Delivery Readiness​

Status: next

Scope:

  • richer release cards
  • delivery-to-planning attachments
  • release blocking model
  • deployment/environment escalation

Deliverables:

  • release readiness card
  • cycle/module release attachment controls
  • environment health summary promoted into Delivery and Recovery

Sprint 3: Git Review and Branch Operations​

Status: planned

Scope:

  • deepen provider-native work

Deliverables:

  • review queue improvements
  • branch action improvements
  • release/review cross-linking
  • provider action summaries in audit

Sprint 4: n8n PM Automation​

Status: planned

Scope:

  • real PM workflow templates

Deliverables:

  • cycle kickoff template
  • release readiness template
  • deployment failure escalation template
  • weekly digest template

Sprint 5: Agent Runtime Bridge​

Status: planned

Scope:

  • make PM work land in actual OpenClaw runtime sessions consistently

Deliverables:

  • mission-control-to-runtime bridge hardening
  • agent dispatch visibility in product UI
  • agent trace timeline per task/handoff

Sprint 6: MCP Productization​

Status: planned

Scope:

  • move MCP from engineering feature to first-class operating surface

Deliverables:

  • documented tool catalog
  • example agent workflows
  • permission model guidance
  • audit-friendly MCP trace views

Sprint 7: Production Readiness Pass​

Status: planned

Scope:

  • performance
  • observability
  • operational maturity

Deliverables:

  • frontend chunking/performance improvements
  • sync error observability
  • retention/backup/runbook docs

Detailed Backlog​

Planning​

Done:

  • cycle CRUD
  • module CRUD
  • page CRUD
  • cycle shared-edit with Plane
  • module shared-edit with Plane
  • page capability fallback

Open:

  • page full shared-edit if Plane API becomes available
  • better conflict resolution UI
  • richer page previews and linking
  • task-to-cycle/module editing improvements in dense boards

Delivery​

Done:

  • pipeline CRUD
  • deployment CRUD
  • provider sync for pipelines/deployments/releases
  • visible pipeline/release sync verification

Open:

  • explicit release blocking/unblocking workflow
  • deployment-to-environment health escalation
  • stronger release/environment quality gates
  • richer release metadata and notes

Work Surface​

Done:

  • planning-aware task metadata
  • cycle/module chips and filters
  • calmer, clearer board UX

Open:

  • more powerful task filtering
  • cycle/module quick assignment from board cards
  • stronger task-history and agent-activity visibility

Risks / Recovery​

Done:

  • stable review/branch/task-follow-up flows
  • live recovery and webhook checks

Open:

  • tighter connection between risky reviews and release readiness
  • better recovery escalations from delivery failures

Settings​

Done:

  • repo/Plane/automation guided journeys
  • visible in-overlay feedback
  • much calmer IA/UX

Open:

  • make advanced/rare integration controls even more compact
  • expose integration capability flags more explicitly

MCP / Agentic Layer​

Done:

  • PM/Git/delivery MCP foundation

Open:

  • richer release/review/deployment tools
  • stronger runtime work dispatch semantics
  • clearer per-agent PM traceability in the product

Acceptance Criteria For “Finished”​

Mission Control can be considered feature-complete for the current vision when all of the following are true:

  • a new project can be created and fully onboarded without manual host intervention
  • planning objects can be operated end-to-end with Plane, or gracefully degrade where unsupported
  • delivery objects reflect real provider state and can influence operator decisions
  • risks, work, planning, delivery, recovery, and automation behave as one coherent suite
  • n8n automation is attached to real PM concepts, not just raw webhooks
  • MCP lets agents operate against PM state safely and audibly
  • the runtime bridge reliably carries Mission Control work into OpenClaw agent execution
  • the full browser smoke suite and live-host suite stay green across releases

Immediate Next Steps​

Recommended immediate order:

  1. finish the Pages product decision and UX:
    • keep graceful-disabled mode if Plane pages remain unsupported
    • or enable full shared-edit once Plane supports it
  2. deepen Delivery:
    • release blocking
    • environment health
    • cycle/module linkage
  3. harden the runtime bridge so Mission Control tasks consistently reach active OpenClaw agents
  4. productize MCP usage and agent workflows
  5. add performance and observability work before broader rollout

References​

  • V2 Architecture (docs/V2_ARCHITECTURE.md)
  • Plane Integration (docs/PLANE_INTEGRATION.md)
  • n8n Operator Integration (docs/N8N_OPERATOR_INTEGRATION.md)
  • Operating Model (docs/OPERATING_MODEL.md)
  • Live Validation (docs/live-validation/ops-signal-orchestra-2026-03-28.md)

License: AGPLv3

Helpifyr: https://helpifyr.com