---
namespace: hm
agent: hm-planner
subtask: true
description: "Cross-AI plan convergence loop — replan with review feedback until no HIGH concerns remain."
argument-hint: "<phase> [--codex] [--gemini] [--claude] [--opencode] [--ollama] [--lm-studio] [--llama-cpp] [--text] [--ws <name>] [--all] [--max-cycles N]"
requires: ["hm-phase", "hm-review"]
validation-gates: ["spec-compliance-gate"]
output-templates: ["hm-plan.md"]
coordination-model: "waiter-model"
completion-signals: ["plan-converged"]
tools:
  read: true
  write: true
  bash: true
  glob: true
  grep: true
  agent: true
  question: true
---


<objective>
Cross-AI plan convergence loop — an outer revision gate around hm-review and hm-planner.
Repeatedly: review plans with external AI CLIs → if HIGH concerns found → replan with --reviews feedback → re-review. Stops when no HIGH concerns remain or max cycles reached.

**Flow:** Agent→Skill("hm-plan-phase") → Agent→Skill("hm-review") → check HIGHs → Agent→Skill("hm-plan-phase --reviews") → Agent→Skill("hm-review") → ... → Converge or escalate

Replaces hm-plan-phase's internal hm-plan-checker with external AI reviewers (codex, gemini, etc.). Each step runs inside an isolated Agent that calls the corresponding existing Skill — orchestrator only does loop control.

**Orchestrator role:** Parse arguments, validate phase, spawn Agents for existing Skills, check HIGHs, stall detection, escalation gate.
</objective>

<execution_context>
@/Users/apple/hivemind-plugin-private/.opencode/workflows/hm-plan-review-convergence.md
@/Users/apple/hivemind-plugin-private/.opencode/references/hm-revision-loop.md
@/Users/apple/hivemind-plugin-private/.opencode/references/hm-gates.md
@/Users/apple/hivemind-plugin-private/.opencode/references/hm-agent-contracts.md
</execution_context>

<runtime_note>
**Copilot (VS Code):** Use `vscode_askquestions` wherever this workflow calls `question`. They are equivalent — `vscode_askquestions` is the VS Code Copilot implementation of the same interactive question API. Do not skip questioning steps because `question` appears unavailable; use `vscode_askquestions` instead.
</runtime_note>

<context>
Phase number: extracted from $ARGUMENTS (required)

**Flags:**
- `--codex` — Use Codex CLI as reviewer (default if no reviewer specified)
- `--gemini` — Use Gemini CLI as reviewer
- `--claude` — Use the agent CLI as reviewer (separate session)
- `--opencode` — Use OpenCode as reviewer
- `--ollama` — Use local Ollama server as reviewer (OpenAI-compatible, default host `http://localhost:11434`; configure model via `review.models.ollama`)
- `--lm-studio` — Use local LM Studio server as reviewer (OpenAI-compatible, default host `http://localhost:1234`; configure model via `review.models.lm_studio`)
- `--llama-cpp` — Use local llama.cpp server as reviewer (OpenAI-compatible, default host `http://localhost:8080`; configure model via `review.models.llama_cpp`)
- `--all` — Use all available CLIs and running local model servers
- `--max-cycles N` — Maximum replan→review cycles (default: 3)

**Feature gate:** This command requires `workflow.plan_review_convergence=true`. Enable with:
`hm config-set workflow.plan_review_convergence true`
</context>

<process>
Execute end-to-end.
Preserve all workflow gates (pre-flight, revision loop, stall detection, escalation).
</process>