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-happy-path.mermaid
View on GitHubstateDiagram-v2
%% Production: context/control_plane.db, human signing key only.
%% Simulation: context/simulation/simulation_control_plane.db, agent test key only.
%% The same lifecycle runs independently; first signed gate binds the context's approver key.
%% Grouping into 4 logical phases matching the lifecycle architecture
state "1. Planning & Spec" as Phase1 {
state "Step 1: Intake — task registered [INTAKE]" as INTAKE
state "Step 2: Interview — baseline and adaptive questions [INTERVIEW]" as INTERVIEW
state "Step 3: Draft plan authored [DRAFT_PLAN]" as DRAFT_PLAN
}
state "2. Plan Review & Gate 1" as Phase2 {
state "Step 4: Plan disposition & acceptance gate [PLAN_REVIEW]" as PLAN_REVIEW
state "Step 5: Independent plan review — 1 or N 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 "3. Execution & Gate 3" as Phase3 {
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 "4. Closure" as Phase4 {
state "Step 12: Retrospective — survey questions [RETROSPECTIVE]" as RETROSPECTIVE
state "Step 13: Done [DONE]" as DONE
}
%% Phase 1 Transitions
[*] --> INTAKE
INTAKE --> INTERVIEW
INTERVIEW --> DRAFT_PLAN : trivial path — short interview<br/>and smallest-change evidence complete
INTERVIEW --> DRAFT_PLAN : standard path — full interview complete
DRAFT_PLAN --> PLAN_REVIEW : submit concise/full plan<br/>trivial skips reviewers by default
%% Phase 2 Transitions & Loops
PLAN_REVIEW --> MULTI_AGENT_REVIEW : optional agent review (0 or more rounds)
PLAN_REVIEW --> AWAITING_APPROVAL : plan accepted by the human (no review required)
PLAN_REVIEW --> DRAFT_PLAN : changes requested<br/>revise artifacts and resubmit
MULTI_AGENT_REVIEW --> PLAN_REVIEW : review complete<br/>merge outcome — repeat or accept
MULTI_AGENT_REVIEW --> DRAFT_PLAN : changes requested<br/>revise artifacts and resubmit
%% CRITICAL HUMAN GATE 1 (cryptographic: signed transition_request)
AWAITING_APPROVAL --> APPROVED : CRITICAL HUMAN GATE 1 (signed) -<br/>human signs the transition_request with<br/>ssh-keygen -Y sign (passphrase prompt)<br/>Releases worktree & coding authority
%% Phase 3 Transitions & Loops
APPROVED --> IN_WORKTREE
IN_WORKTREE --> WORKTREE_REVIEW : implementation + focused tests + inventory done
WORKTREE_REVIEW --> MULTI_AGENT_CODE_REVIEW : you say yes<br/>then choose method/runtime/model/effort
%% CRITICAL HUMAN GATE 3 (cryptographic: signed transition_request, two edges)
WORKTREE_REVIEW --> VERIFY_EXIT : CRITICAL HUMAN GATE 3 (signed) -<br/>human signs a request bound to commit SHA,<br/>tracked-diff hash and untracked-files hash<br/>(direct acceptance, no skip flag exists)
MULTI_AGENT_CODE_REVIEW --> WORKTREE_REVIEW : review package ready for re-submission
MULTI_AGENT_CODE_REVIEW --> IN_WORKTREE : bounded changes<br/>synthesis + human approval
MULTI_AGENT_CODE_REVIEW --> VERIFY_EXIT : CRITICAL HUMAN GATE 3 (signed) -<br/>human signs acceptance of the reviewed code<br/>(same bound request as the direct edge)
%% Phase 4 Transitions
VERIFY_EXIT --> RETROSPECTIVE : deterministic - tests + leak/ledger checks
RETROSPECTIVE --> DONE : CLOSURE GATE (signed) -<br/>human signs a request that also binds<br/>the retrospective digest
DONE --> [*]
%% Bypass & Planning-Only Edges
APPROVED --> RETROSPECTIVE : approved plan — human skips implementation
INTERVIEW --> RETROSPECTIVE : emergency trivial/early close only
INTAKE --> DONE : early close — same signed closure gate<br/>(any active state, no force-close string or flag)
note right of APPROVED
Critical Gate 1 (cryptographic):
Unlocks file write authority and isolated
worktree execution. Only a verified OpenSSH
signature over the transition_request advances it.
end note
note right of WORKTREE_REVIEW
Critical Gate 3 (cryptographic):
Both edges into VERIFY_EXIT (direct and post-review)
require a human signature over commit SHA + diff hash
+ untracked-files hash. The signature is verified and
the request consumed once, in the commit transaction.
Receipts, typed words, flags and actor strings cannot
substitute.
end note
note right of DONE
Every edge into DONE (normal closure and early close)
needs the same human-signed transition_request.
Production pushes reverify complete signed gate history.
Simulation DONE cannot authorize a production push.
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 | Unverified | 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 |