mma

Plugin mma from marketplace multi-model-agent in zhixuan312/multi-model-agent. Commands read from the repo's manifest — 1 month ago.

Skills (17)

• audit

  Use when the user asks to audit a spec / plan / design doc / skill file. The `subtype` field picks the criteria set. `default` (prose-coherence) is the general doc auditor. `plan` verifies a code-execution plan against the actual codebase — run this before any `mma:execute-plan` dispatch. `spec` aud

• brainstorm

  Use when a grounded idea needs to be grilled into a formal spec — runs a relentless one-decision-at-a-time requirement interview, resolving mechanical questions via workers and putting only real decisions to the user, then dispatches mma:spec. Not for grounding a raw idea (use mma:explore first) or

• context-blocks

  Use when a document larger than ~2 KB will be referenced by 2+ subsequent mma:* calls — register once, pass the returned ID to each call instead of re-uploading the same content. OR a spec / plan / error log was already inlined into one task and is about to be inlined into a second — register on the

• debug

  Use when a deliverable is wrong — a test fails, a build breaks, a report shows the wrong figures, a workflow misbehaves — AND narrowing the root cause requires reading files, reproducing the failure, or tracing across multiple modules — the worker investigates so the main agent stays on the hypothes

• delegate

  Use when you have an ad-hoc implementation or research task WITHOUT a plan file on disk and you want it to run on a cheap worker instead of consuming main-context tokens

• execute-plan

  Use when a plan or spec file exists on disk (any markdown with task headings — .mma/plans/*.md, a TODO list, a spec doc) and you need to implement one or more tasks from it sequentially in one worker session

• explore

  Use when a raw idea or problem needs grounding before it's designed — captures a braindump, fans out divergent internal-codebase investigation + external research + prior-learnings recall in parallel, then synthesises the results into a written exploration.md (Background · Current state · Rough dire

• investigate

  Use when you need to answer a question about the project ("how does X work", "where is Y called", "what does this directory do") and reading + grepping the material yourself would consume main-context tokens

• journal-recall

  Use when you're about to design or attempt something and want to know what THIS project already learned — ask a vague conceptual question (no tags or keywords needed); a read-only worker searches the learnings graph and returns the relevant prior lessons + how they relate. Fire before re-treading gr

• journal-record

  Use when you've learned something worth remembering — a decision, design rationale, user behavior pattern, process learning, research finding, or style convention. Records it to the persistent team knowledge graph for future sessions.

• orchestrate

  Use when a frontend workflow needs a high-quality LLM brain for orchestration — send a structured prompt, get a structured response, reuse the session across workflow phases

• plan

  Use when you have a spec file on disk and need a contract-first implementation plan written by a worker — produces ordered Contract Tasks (inputs/outputs/mapping/errors/invariants) with plan-authored acceptance tests and no implementation code, plus exact file paths

• research

  Use when you need external multi-source research with citations — arxiv, semantic_scholar, github_search, brave-with-site:-filters — for a focused question. Worker is bibliographic, not opinionated. Pair with mma:investigate (internal) under mma:explore for divergent landscape scans.

• review

  Use when source code needs a quality / security / correctness pass — pre-merge review, post-implementation sanity check, or focused look at a small file set — over a small file set the worker reads together

• router

  Use first whenever you're about to delegate any tool-using work — picks the right mma:* skill (audit, review, debug, plan execution, codebase investigation, ad-hoc delegation, context-block reuse) instead of defaulting to inline Agent dispatches

• solution-lead

  Use when a business stakeholder describes a goal in their own words and it needs to be run end to end — understand the goal, draft and confirm the initiative in plain language, create the durable record only after confirmation, coordinate the work, and report back with real verification evidence.

• spec

  Use when you have structured design decisions ready and need a formal specification document written by a worker instead of consuming main-context tokens

Slash commands (4)

• /mma:breakout

  Claude Code command: /mma:breakout — spawn a named expert breakout teammate, keep the deep dialogue isolated, and close with one confirmed journal batch

• /mma:deck

  Claude Code command: /mma:deck — turn a source document into a slide deck built on the house visual system. Find the argument in the source, choose a composition per claim, and emit one standalone offline HTML file.

• /mma:flow

  Claude Code command: /mma:flow — MMA-native solution delivery playbook (pr, commit-in-place, or deliver-file) that resumes from durable artifacts

• /mma:tldr

  Claude Code command: /mma:tldr — turn the previous assistant message or a supplied source into a short decision brief. Compress long material, explain difficult material, keep material qualifications, and name omitted topics.

MCP servers (1)

• daemon

  connects to: http://127.0.0.1:7337/mcp

Similar plugins