richfrem/agent-plugins-skills / plugins/agent-agentic-os/.apm/skills/running-control-plane
running-control-plane
FlaggedCRITICAL INTAKE GATEWAY: use at the start of any non-trivial engineering task, feature, refactor, multi-file bugfix, work package or new idea, including requests to explore, brainstorm, build, fix or change something, before planning or code changes. Enforces read-only discovery, 1-3 structured questions with defaults, control-plane registration and the immutable four-pillar TASK_SPEC.md. Also runs the rest of the control plane: health ("run health check", "check os metrics", "system monitor", event bus, locks, memory, "is my signing identity ready" status check), transition simulation ("dry run transition", "simulate transition", "check edge X -> Y"), worktree selection, and gating on/off ("disable the control plane", "turn off the push guard", "stop gating my commits"). Not for project setup or key creation (use setting-up-agentic-os) or improving skills (use improving-skills-in-lab).
references/diagrams/control-plane-pipeline.mermaid
View on GitHubstateDiagram-v2
%% This drawing names the production human gates. Simulations run the same lifecycle
%% independently in context/simulation/simulation_control_plane.db with the agent key only.
%% First signed gate binds one key; simulation evidence never authorizes production pushes.
state "Step 1: Intake — task registered [INTAKE]" as INTAKE
state "Step 2: Interview — understanding what you want [INTERVIEW]" as INTERVIEW
state "Step 3: Draft plan authored [DRAFT_PLAN]" as DRAFT_PLAN
state "Step 4: Plan disposition & acceptance gate [PLAN_REVIEW]" as PLAN_REVIEW
state "Step 5: Independent plan review — 0 or more rounds [MULTI_AGENT_REVIEW]" as MULTI_AGENT_REVIEW
state "Step 6: Awaiting your approval [AWAITING_APPROVAL]" as AWAITING_APPROVAL
state "Step 7: Approved — you said go [APPROVED]" as APPROVED
state "Step 8: Implementation (isolated worktree) [IN_WORKTREE]" as IN_WORKTREE
state "Step 9: Here's what changed — initial review [WORKTREE_REVIEW]" as WORKTREE_REVIEW
state "Step 10: Multi-agent code review — 1 or N rounds [MULTI_AGENT_CODE_REVIEW]" as MULTI_AGENT_CODE_REVIEW
state "Step 11: Verifying (tests, checks) [VERIFY_EXIT]" as VERIFY_EXIT
state "Step 12: Retrospective — survey questions [RETROSPECTIVE]" as RETROSPECTIVE
state "Step 13: Done [DONE]" as DONE
state "Rolled back — problem found, documented [ROLLED_BACK]" as ROLLED_BACK
state "Escalated to you for a decision [ESCALATED]" as ESCALATED
[*] --> INTAKE
INTAKE --> INTERVIEW
INTAKE --> DRAFT_PLAN : skip straight to planning
INTAKE --> PLAN_REVIEW : skip straight to planning
INTAKE --> ESCALATED
INTAKE --> DONE : CLOSURE GATE — human-signed transition_request
INTERVIEW --> DRAFT_PLAN : interview complete<br/>no plan artifacts required yet
INTERVIEW --> PLAN_REVIEW : legacy/direct plan submission path<br/>(artifacts required)
INTERVIEW --> RETROSPECTIVE : trivial path — interview baseline<br/>and adaptive follow-ups complete
INTERVIEW --> ESCALATED
INTERVIEW --> INTAKE : human-approved recovery reset
INTERVIEW --> DONE : CLOSURE GATE — human-signed transition_request
DRAFT_PLAN --> PLAN_REVIEW : submit plan<br/>ask additional review? yes/no
PLAN_REVIEW --> DRAFT_PLAN : changes requested<br/>revise artifact identity
DRAFT_PLAN --> INTERVIEW : needs more questions
PLAN_REVIEW --> INTERVIEW : needs more questions
PLAN_REVIEW --> MULTI_AGENT_REVIEW : optional agent review (0 or more rounds)<br/>ask method, then internal runtime/model/effort
PLAN_REVIEW --> AWAITING_APPROVAL : plan accepted by the human (no review required)<br/>continue to human approval
DRAFT_PLAN --> ESCALATED
DRAFT_PLAN --> INTAKE : human-approved recovery reset
DRAFT_PLAN --> DONE : CLOSURE GATE — human-signed transition_request
PLAN_REVIEW --> ESCALATED
PLAN_REVIEW --> INTAKE : human-approved recovery reset
PLAN_REVIEW --> DONE : CLOSURE GATE — human-signed transition_request
MULTI_AGENT_REVIEW --> PLAN_REVIEW : review complete<br/>merge outcomes for plan acceptance
MULTI_AGENT_REVIEW --> DRAFT_PLAN : reviewer requested changes<br/>new identity required before resubmission
MULTI_AGENT_REVIEW --> ESCALATED
MULTI_AGENT_REVIEW --> INTAKE : human-approved recovery reset
MULTI_AGENT_REVIEW --> DONE : CLOSURE GATE — human-signed transition_request
%% CRITICAL HUMAN GATE 1 (cryptographic)
AWAITING_APPROVAL --> APPROVED : CRITICAL HUMAN GATE 1 (signed) -<br/>human signs the transition_request<br/>(ssh-keygen -Y sign, passphrase prompt)<br/>Releases worktree & coding authority
AWAITING_APPROVAL --> DRAFT_PLAN : needs more work
AWAITING_APPROVAL --> PLAN_REVIEW : needs more work
AWAITING_APPROVAL --> ESCALATED
AWAITING_APPROVAL --> INTAKE : human-approved recovery reset
AWAITING_APPROVAL --> DONE : CLOSURE GATE — human-signed transition_request
APPROVED --> IN_WORKTREE
APPROVED --> RETROSPECTIVE : human-approved skip implementation
APPROVED --> ESCALATED
APPROVED --> INTAKE : human-approved recovery reset
APPROVED --> DONE : CLOSURE GATE — human-signed transition_request
IN_WORKTREE --> WORKTREE_REVIEW : implementation + tests done
IN_WORKTREE --> ROLLED_BACK
IN_WORKTREE --> ESCALATED
IN_WORKTREE --> INTAKE : human-approved recovery reset
IN_WORKTREE --> DONE : CLOSURE GATE — human-signed transition_request
WORKTREE_REVIEW --> MULTI_AGENT_CODE_REVIEW : yes — choose method<br/>then internal runtime/model/effort one at a time
%% CRITICAL HUMAN GATE 3b (cryptographic)
WORKTREE_REVIEW --> VERIFY_EXIT : CRITICAL HUMAN GATE 3b (signed) -<br/>human signs acceptance bound to commit SHA,<br/>diff hash and untracked-files hash
WORKTREE_REVIEW --> IN_WORKTREE : more changes needed
WORKTREE_REVIEW --> ROLLED_BACK
WORKTREE_REVIEW --> ESCALATED
WORKTREE_REVIEW --> INTAKE : human-approved recovery reset
WORKTREE_REVIEW --> DONE : CLOSURE GATE — human-signed transition_request
MULTI_AGENT_CODE_REVIEW --> WORKTREE_REVIEW : review package ready<br/>for artifact/scope confirmation
%% CRITICAL HUMAN GATE 3c (cryptographic)
MULTI_AGENT_CODE_REVIEW --> VERIFY_EXIT : CRITICAL HUMAN GATE 3c (signed) -<br/>human signs acceptance of the reviewed code<br/>(same bound request as 3b)
MULTI_AGENT_CODE_REVIEW --> IN_WORKTREE : synthesis says bounded changes<br/>human approves rework
MULTI_AGENT_CODE_REVIEW --> ROLLED_BACK
MULTI_AGENT_CODE_REVIEW --> ESCALATED
MULTI_AGENT_CODE_REVIEW --> INTAKE : human-approved recovery reset
MULTI_AGENT_CODE_REVIEW --> DONE : CLOSURE GATE — human-signed transition_request
VERIFY_EXIT --> RETROSPECTIVE : ledger + full/task tests + leak check<br/>and map-debt evidence pass
VERIFY_EXIT --> IN_WORKTREE : verification failed
VERIFY_EXIT --> WORKTREE_REVIEW : verification failed
VERIFY_EXIT --> ROLLED_BACK
VERIFY_EXIT --> ESCALATED
VERIFY_EXIT --> INTAKE : human-approved recovery reset
VERIFY_EXIT --> DONE : CLOSURE GATE — human-signed transition_request
RETROSPECTIVE --> DONE : CLOSURE GATE — human-signed transition_request<br/>binds the retrospective digest<br/>follow-ups resolved
RETROSPECTIVE --> ESCALATED : retrospective blocked
RETROSPECTIVE --> INTAKE : human-approved recovery reset
ROLLED_BACK --> ESCALATED
ROLLED_BACK --> PLAN_REVIEW : restart planning
ROLLED_BACK --> INTAKE : human-approved recovery reset
ROLLED_BACK --> DONE : CLOSURE GATE — human-signed transition_request
ROLLED_BACK --> RETROSPECTIVE : human-approved recovery reset
ESCALATED --> INTAKE : start over
ESCALATED --> PLAN_REVIEW : restart planning
ESCALATED --> DONE : CLOSURE GATE — human-signed transition_request
ESCALATED --> RETROSPECTIVE : human-approved recovery reset
DONE --> INTAKE : human-approved recovery reset
DONE --> RETROSPECTIVE : human-approved recovery reset
INTAKE --> RETROSPECTIVE : human-approved recovery reset
DRAFT_PLAN --> RETROSPECTIVE : human-approved recovery reset
MULTI_AGENT_REVIEW --> RETROSPECTIVE : human-approved recovery reset
PLAN_REVIEW --> RETROSPECTIVE : human-approved recovery reset
AWAITING_APPROVAL --> RETROSPECTIVE : human-approved recovery reset
IN_WORKTREE --> RETROSPECTIVE : human-approved recovery reset
WORKTREE_REVIEW --> RETROSPECTIVE : human-approved recovery reset
MULTI_AGENT_CODE_REVIEW --> RETROSPECTIVE : human-approved recovery reset
DONE --> [*]
note right of APPROVED
Critical Gate 1 (cryptographic):
Unlocks file write authority and isolated
worktree execution. Only a verified OpenSSH
signature (ssh-keygen -Y sign) over the
transition_request advances this edge.
end note
note right of WORKTREE_REVIEW
Critical Gate 3b & 3c (cryptographic):
Entering VERIFY_EXIT requires a human signature over
commit SHA + tracked-diff hash + untracked-files hash.
The request is verified and consumed once, inside the
commit transaction. No skip flag, receipt, typed word
or actor string can substitute; there is no SQLite
trigger that trusts an actor column for this gate.
end note
note right of DONE
Every edge into DONE -- normal closure and early
close from any active state -- requires the same
human-signed transition_request. There is no
force-close string, flag or confirmation question.
end note
note right of IN_WORKTREE
After APPROVED → IN_WORKTREE, the
implementation controller runs dispatch,
review, synthesis, and bounded fixes without
asking for routine lifecycle transitions.
Re-enter review only after focused tests and
the session-wide change inventory pass.
No direct IN_WORKTREE -> VERIFY_EXIT edge exists.
A "Gate 3d" fix once gated that edge with a human
question instead of removing it; the edge was removed
entirely (2026-09-17): all work passes through
WORKTREE_REVIEW first.
end note
note right of MULTI_AGENT_CODE_REVIEW
Plan-review receipts and implementation-review
receipts are isolated. Every reviewer report is
preserved, synthesized into one revision brief,
and a requested change returns through the
human-approved IN_WORKTREE rework edge.
end note
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |