Two different programs use “Phase 0” naming in Perpetua-Tools. They are not the same workstream. Link here whenever a doc says “Phase 0” without a folder path.
| If you see… | It means… |
|---|---|
Phase 0 (no letter) + phase-0-specifications/ |
STM / swarm security — PeerObservation, STM, T1–T7, mesh evidence |
| Phase 0F | Coordination CLI freeze — lock agent_coordination.py’s 29 commands before more extraction |
0F = “freeze” inside the coordination consolidation migration ladder
(0F → Part 1 → Part 2…). It is not “Phase 0 subsection F” of the STM
graph.
| Question | How do peer observations, liveness, replay, quorum, and threats become one decision boundary? |
| Canonical home | ../phase-0-specifications/README.md + ../phase-0-specifications/wiki/ |
| Master checklist | ../phase-0-specifications/PHASE-0-MASTER-PLAN-2026-07-27.md |
| Runtime | orchestrator/state_transition_manager.py, membership.py (tested; dormant in production per descope) |
| “Phase 1 / 2” here | STM integration phases (1.0–1.3, job board) — security/STM, not coordination |
| Orama tie-in | D23 descope, mesh finality, fleet-mesh security companions |
Use when discussing: T1–T7, evaluate_observation(), descope of P5/P6/P13,
PR #203/#205, swarm security analysis.
| Question | What is the live CLI/parser/dispatch contract before more code moves into orchestrator/coordination/? |
| Canonical home | README.md (this folder) |
| Mother plan | ../next/2026-07-17-coordination-module-consolidation-plan.md |
| Review (in-repo) | ../references/coordination-consolidation-plan-review-2026-07-18.md — supersedes off-repo handoff copies |
| Runtime | orchestrator/coordination/cli.py, scripts/agent_coordination.py |
| 0F artifact | tests/test_agent_coordination_cli_contract.py (29 leaves + handler dispatch) |
| Part 2 after 0F | liveness.py extraction, module split completion, compat deletion gate |
Use when discussing: queue claim races, heartbeat handlers, make_gossip_bus vs
GossipBus, provenance table, “freeze 29 CLI leaves.”
STM Phase 0 (security) Coordination Phase 0F (CLI freeze)
───────────────────── ─────────────────────────────────
PeerObservation / STM agent_coordination 29 commands
Threat T1–T7 Parser + dispatch contract
docs/phase-0-specifications/ docs/coordination/
DORMANT BFT in prod Part 2: liveness.py extraction
Mesh P5/P6 (orama runtime) ≠ STM P5/P6 (dormant on PT)
Mesh P5/P6 (swarm approval, discovery trust on orama) and STM P5/P6/P13 (witness/reputation on PT) share pattern names but live in different repos and layers.
| Term | Domain |
|---|---|
| gbrain CRG Phase 0–1 | Code-review-graph indexing — unrelated |
| Coordination “Phase 1” skeleton | orchestrator/coordination/ package layout — coordination ladder, not STM Phase 1 |
| Fleet Phase 7–10+ | GossipBus / G7 — separate numbering line |
When a doc says “Phase 0” without a path, check the folder
(phase-0-specifications vs coordination vs fleet-mesh) or the letter F
(0F).
| Avoid | Prefer |
|---|---|
| “Phase 0” alone | STM Phase 0 or coordination Phase 0F |
| “Phase 0F” in STM docs | Never — keep STM graph free of 0F |
| “Phase 0 master plan” for coordination | Coordination CLI contract (Phase 0F) |