Skip to content
Back to search

mcode-stack

Workflows and skills the MiniMax Code team uses to build MiniMax Code. Each stack installs as one plugin.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/minimax-ai--mcode-stack/badge.svg)](/listings/minimax-ai--mcode-stack)

HTML

<a href="/listings/minimax-ai--mcode-stack"><img src="/api/listings/minimax-ai--mcode-stack/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

**Claude Code, one command.** The command adds this marketplace and installs a whole stack. In a shell:

claude plugin install ronny --marketplace MiniMax-AI/mcode-stack

Inside a Claude Code session:

/plugin install ronny --marketplace MiniMax-AI/mcode-stack

Use `jiachen`, `req-tpm` or `youfang` in place of `ronny` for those stacks. From a clone of this repository, `scripts/install.sh` installs `ronny` (`scripts/install.sh jiachen`, `scripts/install.sh req-tpm` or `scripts/install.sh youfang` for the others) with `claude` or `codex` when either is on `PATH`, and otherwise copies that stack's skills into `~/.agents/skills/`.

Older Claude Code can add the marketplace first, then install:

/plugin marketplace add MiniMax-AI/mcode-stack
/plugin install ronny@mcode-stack

**Codex.** Add the marketplace, then install from the Plugins Directory or the CLI:

codex plugin marketplace add MiniMax-AI/mcode-stack
codex plugin add ronny@mcode-stack

Substitute `jiachen`, `req-tpm` or `youfang` the same way. To take a single skill instead of the plugin, copy that skill's folder into `~/.agents/skills/`. Codex and MiniMax Code both scan that directory; the copy steps are below.

**One skill, for all your projects:**

git clone https://github.com/MiniMax-AI/mcode-stack.git ~/mcode-stack
mkdir -p ~/.agents/skills
cp -R ~/mcode-stack/stacks/ronny/skills/long-command-check ~/.agents/skills/

Open a new session and type `/skills` in MiniMax Code to see it. Run it with `/long-command-check`, or let the agent load it from its description.

**The whole stack, as a plugin:**

mcode plugin marketplace list    # shows your local plugin directory

Copy the whole `stacks/ronny` directory into that directory, then open `/plugins` in MiniMax Code and press `Ctrl+R` to refresh. A local plugin is installed as soon as it is in that directory; press Space in the panel to enable or disable it.

**Every stack, from the MiniMax Code marketplace:**

This repository is also a plugin marketplace, so you can install a stack without copying files and pick up updates later. In MiniMax Code desktop, open **Plugins**, choose **Add marketplace** and enter `MiniMax-AI/mcode-stack`; each stack then shows up as one plugin you can install, update or uninstall. Adding marketplaces is rolling out gradually, so your version may not have it yet.

<details> <summary>More install options</summary>

MiniMax Code discovers skills in these directories (one subdirectory per skill, containing `SKILL.md`):

- For the current project only: `.minimax/skills/`, `.claude/skills/` or `.agents/skills/` in the project - For all projects: `~/.agents/skills/` (`~/.claude/skills/` and `~/.codex/skills/` are read too)

`mcode plugin add` installs from the official catalog only; it does not take a GitHub URL or a local path.

The skills use the common `SKILL.md` format, so most of them also work in other agents that support skills.

</details>

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 51 lines
  1. {
  2. "$schema": "https://json.schemastore.org/claude-code-marketplace.json",
  3. "name": "mcode-stack",
  4. "description": "Workflows and skills the MiniMax Code team uses to build MiniMax Code. Each stack installs as one plugin.",
  5. "owner": {
  6. "name": "MiniMax Code team",
  7. "url": "https://github.com/MiniMax-AI/mcode-stack"
  8. },
  9. "plugins": [
  10. {
  11. "name": "ronny",
  12. "source": "./stacks/ronny",
  13. "description": "Daily workflows and self-check skills from the MiniMax Code R&D lead: independent review, subagent split checks, tool and context budget audits, long sessions, long commands, harness experiments, trajectory review, agent briefs, evidence-first review of agent output, review capacity and turning repeat corrections into checks, plus everyday engineering practice for code review, review comments, small PRs, change descriptions and emergency changes.",
  14. "category": "Code",
  15. "author": {
  16. "name": "Ronny (MiniMax Code team)"
  17. },
  18. "homepage": "https://github.com/MiniMax-AI/mcode-stack/tree/main/stacks/ronny"
  19. },
  20. {
  21. "name": "jiachen",
  22. "source": "./stacks/jiachen",
  23. "description": "MCode Review, PR review that has to prove its findings: the same rules in your agent and in GitHub or GitLab CI, a verdict decided by a script, threads that carry over between pushes, and a weekly loop that tunes the rules on your own history. Plus finding the loops where you keep acting as a manual step for your agents, and a content pipeline that turns material into Xiaohongshu image posts and short vertical videos.",
  24. "category": "Code",
  25. "author": {
  26. "name": "Jiachen (MiniMax Code team)"
  27. },
  28. "homepage": "https://github.com/yujiachen-y/dotfiles"
  29. },
  30. {
  31. "name": "req-tpm",
  32. "source": "./stacks/req-tpm",
  33. "description": "Requirement management where the requester and their agent act as the requirement's TPM: intake with a measurable expected outcome, priority recommendation for the owner, design decision record, cadence tracking that flags stale items, ship readiness, and evidence-based outcome review. Body in a Feishu doc, status in Meegle.",
  34. "category": "Productivity",
  35. "author": {
  36. "name": "MiniMax Code team"
  37. },
  38. "homepage": "https://github.com/MiniMax-AI/mcode-stack/tree/main/stacks/req-tpm"
  39. },
  40. {
  41. "name": "youfang",
  42. "source": "./stacks/youfang",
  43. "description": "Turn architecture and flow diagrams into Keynote-style Magic Move slide decks: one offline HTML file where the boxes and titles two slides share morph and fly, in one of four visual styles (hand-drawn, minimal, dark neon, neo-brutalist) picked from rendered samples, with optional 4K video export.",
  44. "category": "Productivity",
  45. "author": {
  46. "name": "Youfang (MiniMax Code team)"
  47. },
  48. "homepage": "https://github.com/MiniMax-AI/mcode-stack/tree/main/stacks/youfang"
  49. }
  50. ]
  51. }
README.md · README · 264 lines
  1. <p align="center">
  2. <img src="docs/assets/hero.png" alt="MCode Stack: the workflows and skills the MiniMax Code team uses to build MiniMax Code" width="100%">
  3. </p>
  4. <p align="center">
  5. <a href="LICENSE"><img alt="License: MIT" src="https://img.shields.io/badge/license-MIT-blue.svg"></a>
  6. <a href="https://github.com/MiniMax-AI/mcode-stack/actions/workflows/validate.yml"><img alt="validate" src="https://github.com/MiniMax-AI/mcode-stack/actions/workflows/validate.yml/badge.svg"></a>
  7. <a href="stacks/"><img alt="skills: stacks/" src="https://img.shields.io/badge/skills-stacks%2F-ff7a59.svg"></a>
  8. <a href="https://github.com/MiniMax-AI/minimax-code"><img alt="for MiniMax Code" src="https://img.shields.io/badge/for-MiniMax%20Code-5b6cff.svg"></a>
  9. </p>
  10. <p align="center"><b>English</b> · <a href="README_ZH.md">中文</a></p>
  11. **The workflows and skills the MiniMax Code team uses to build MiniMax Code. Open source, install and use.**
  12. This is not a prompt collection or a list of "best practices". It is what
  13. people on the team actually use every day: how to keep AI-written code at a
  14. high bar, how to decide whether a task should go to a subagent, how long to
  15. wait on a long command, how to keep long sessions from going dull, and what
  16. goes into context on every turn. Each piece is a `SKILL.md` skill or
  17. checklist that installs straight into
  18. [MiniMax Code](https://github.com/MiniMax-AI/minimax-code).
  19. - [Quick start](#quick-start)
  20. - [What's inside](#whats-inside)
  21. - [How it works](#how-it-works)
  22. - [Articles](#articles)
  23. - [Stacks](#stacks)
  24. - [Contributing](#contributing)
  25. ## Quick start
  26. **Claude Code, one command.** The command adds this marketplace and installs
  27. a whole stack. In a shell:
  28. ```bash
  29. claude plugin install ronny --marketplace MiniMax-AI/mcode-stack
  30. ```
  31. Inside a Claude Code session:
  32. ```bash
  33. /plugin install ronny --marketplace MiniMax-AI/mcode-stack
  34. ```
  35. Use `jiachen`, `req-tpm` or `youfang` in place of `ronny` for those stacks. From a clone of this
  36. repository, `scripts/install.sh` installs `ronny` (`scripts/install.sh jiachen`,
  37. `scripts/install.sh req-tpm` or `scripts/install.sh youfang` for the others) with `claude` or `codex` when either is on `PATH`, and
  38. otherwise copies that stack's skills into `~/.agents/skills/`.
  39. Older Claude Code can add the marketplace first, then install:
  40. ```bash
  41. /plugin marketplace add MiniMax-AI/mcode-stack
  42. /plugin install ronny@mcode-stack
  43. ```
  44. **Codex.** Add the marketplace, then install from the Plugins Directory or the
  45. CLI:
  46. ```bash
  47. codex plugin marketplace add MiniMax-AI/mcode-stack
  48. codex plugin add ronny@mcode-stack
  49. ```
  50. Substitute `jiachen`, `req-tpm` or `youfang` the same way. To take a single skill instead of the
  51. plugin, copy that skill's folder into `~/.agents/skills/`. Codex and MiniMax
  52. Code both scan that directory; the copy steps are below.
  53. **One skill, for all your projects:**
  54. ```bash
  55. git clone https://github.com/MiniMax-AI/mcode-stack.git ~/mcode-stack
  56. mkdir -p ~/.agents/skills
  57. cp -R ~/mcode-stack/stacks/ronny/skills/long-command-check ~/.agents/skills/
  58. ```
  59. Open a new session and type `/skills` in MiniMax Code to see it. Run it with
  60. `/long-command-check`, or let the agent load it from its description.
  61. **The whole stack, as a plugin:**
  62. ```bash
  63. mcode plugin marketplace list # shows your local plugin directory
  64. ```
  65. Copy the whole `stacks/ronny` directory into that directory, then open
  66. `/plugins` in MiniMax Code and press `Ctrl+R` to refresh. A local plugin is
  67. installed as soon as it is in that directory; press Space in the panel to
  68. enable or disable it.
  69. **Every stack, from the MiniMax Code marketplace:**
  70. This repository is also a plugin marketplace, so you can install a stack
  71. without copying files and pick up updates later. In MiniMax Code desktop, open
  72. **Plugins**, choose **Add marketplace** and enter `MiniMax-AI/mcode-stack`; each
  73. stack then shows up as one plugin you can install, update or uninstall. Adding
  74. marketplaces is rolling out gradually, so your version may not have it yet.
  75. <details>
  76. <summary>More install options</summary>
  77. MiniMax Code discovers skills in these directories (one subdirectory per
  78. skill, containing `SKILL.md`):
  79. - For the current project only: `.minimax/skills/`, `.claude/skills/` or
  80. `.agents/skills/` in the project
  81. - For all projects: `~/.agents/skills/` (`~/.claude/skills/` and
  82. `~/.codex/skills/` are read too)
  83. `mcode plugin add` installs from the official catalog only; it does not take
  84. a GitHub URL or a local path.
  85. The skills use the common `SKILL.md` format, so most of them also work in
  86. other agents that support skills.
  87. </details>
  88. ## What's inside
  89. <p align="center">
  90. <img src="docs/assets/skills-map.png" alt="The skills in stacks/ronny, grouped by moment in the day, plus what stacks/jiachen adds" width="100%">
  91. </p>
  92. | When | Skill | What it does |
  93. | --- | --- | --- |
  94. | Plan & split | [`subagent-split-check`](stacks/ronny/skills/subagent-split-check/SKILL.md) | Decide whether work goes to a subagent or stays in the main loop |
  95. | | [`engineering-small-prs`](stacks/ronny/skills/engineering-small-prs/SKILL.md) | Split large features, refactors and migrations into small reviewable PRs |
  96. | | [`engineering-change-descriptions`](stacks/ronny/skills/engineering-change-descriptions/SKILL.md) | Write PR and commit descriptions that say what changed and why |
  97. | Context & tools | [`tool-budget-audit`](stacks/ronny/skills/tool-budget-audit/SKILL.md) | Measure what tools cost in context; trim, tool search or programmatic tool calling |
  98. | | [`context-tax-audit`](stacks/ronny/skills/context-tax-audit/SKILL.md) | What gets injected into every request, and what it does to the prompt cache |
  99. | Long sessions & commands | [`long-session-check`](stacks/ronny/skills/long-session-check/SKILL.md) | Drop, summarize or write to file; compact or start fresh; memory scoping |
  100. | | [`handoff-doc`](stacks/ronny/skills/handoff-doc/SKILL.md) | A hand-off doc so a new session can pick up the work |
  101. | | [`long-command-check`](stacks/ronny/skills/long-command-check/SKILL.md) | Ten-point audit of long shell commands and background tasks |
  102. | Review | [`independent-review`](stacks/ronny/skills/independent-review/SKILL.md) | Review a diff from a clean context: only diff, requirement and test results |
  103. | | [`engineering-code-review`](stacks/ronny/skills/engineering-code-review/SKILL.md) | Review a change for code health: design, functionality, complexity, tests, approval risk |
  104. | | [`engineering-review-comments`](stacks/ronny/skills/engineering-review-comments/SKILL.md) | Write clear, severity-labeled review comments that explain the reasoning |
  105. | | [`engineering-review-feedback`](stacks/ronny/skills/engineering-review-feedback/SKILL.md) | Handle reviewer comments as the author: fix, clarify, resolve disagreements |
  106. | Experiments & hotfixes | [`harness-experiment`](stacks/ronny/skills/harness-experiment/SKILL.md) | Small eval, one change at a time, repeated runs, look at flipped tasks |
  107. | | [`trajectory-review`](stacks/ronny/skills/trajectory-review/SKILL.md) | Look at how the agent got its result, not only whether it passed |
  108. | Owning agent work | [`agent-brief`](stacks/ronny/skills/agent-brief/SKILL.md) | Before delegating: stakes, hypothesis, scope, what "verified" means |
  109. | | [`evidence-first-review`](stacks/ronny/skills/evidence-first-review/SKILL.md) | When an agent says done: evidence first, what checks miss, silent decisions |
  110. | | [`review-capacity-check`](stacks/ronny/skills/review-capacity-check/SKILL.md) | Don't launch more agents than we can stand behind |
  111. | | [`taste-to-checks`](stacks/ronny/skills/taste-to-checks/SKILL.md) | Turn a repeat correction into a test, lint, hook or CI step |
  112. | | [`engineering-emergency-changes`](stacks/ronny/skills/engineering-emergency-changes/SKILL.md) | Decide whether a hotfix is a real emergency and run expedited review safely |
  113. The table above is [`stacks/ronny`](stacks/ronny/).
  114. [`stacks/jiachen`](stacks/jiachen/) adds **MCode Review**, PR review that has
  115. to prove its findings: the same rules in your agent
  116. ([`pr-review`](stacks/jiachen/skills/pr-review/SKILL.md)) and in GitHub or
  117. GitLab CI, a verdict decided by a script, and a weekly loop
  118. ([`pr-review-tuning`](stacks/jiachen/skills/pr-review-tuning/SKILL.md)) that
  119. tunes its rules on your own PRs. It also adds
  120. [`lifting-loops`](stacks/jiachen/skills/lifting-loops/SKILL.md), which finds
  121. the loops where you keep acting as a manual step for your agents and lifts you
  122. out of them, and [`making-content`](stacks/jiachen/skills/making-content/SKILL.md),
  123. which turns notes and material into Xiaohongshu image posts and short vertical
  124. videos.
  125. [`stacks/req-tpm`](stacks/req-tpm/) is an optional stack for requirement
  126. management. The person who raises a product or technical requirement acts as
  127. its TPM together with their own agent: challenge it at intake and fix a
  128. measurable expected outcome, get a priority decision from the owner, record
  129. the design decision, track it on a cadence and flag stale items, pass ship
  130. gates, and close only with evidence against the expected outcome. The body
  131. lives in a Feishu doc; status and flow live in Meegle (Feishu Project). It is
  132. v0 and needs the `meegle` and `lark-cli` CLIs plus a one-time look-up of your
  133. Meegle space's fields and states; see
  134. [`stacks/req-tpm/README.md`](stacks/req-tpm/README.md).
  135. [`stacks/youfang`](stacks/youfang/) has
  136. [`magic-move-deck`](stacks/youfang/skills/magic-move-deck/SKILL.md), which
  137. turns an architecture or flow walkthrough into a Keynote-style Magic Move
  138. slide deck: every slide is a diagram seen through a camera, and the boxes and
  139. titles two slides share morph between them. The agent shows one sample slide
  140. in each of four styles (hand-drawn, minimal, dark neon, neo-brutalist) and
  141. builds the deck in the one you pick, as a single offline HTML file.
  142. A stack usually has three kinds of things:
  143. | Kind | What it is | How to use it |
  144. | --- | --- | --- |
  145. | Skill | A directory with a `SKILL.md` | Invoke it in MiniMax Code with `/<skill-name>`, or let the agent load it from its description |
  146. | Checklist | Also a skill: the agent walks your project item by item. A plain-text copy lives in `checklists/` | Same as above, or just read it |
  147. | Workflow | A `WORKFLOW.md` describing how this person works with MiniMax Code every day | Read it |
  148. ## How it works
  149. <p align="center">
  150. <img src="docs/assets/how-it-works.png" alt="From an Engineering Notes article to a skill, installed into MiniMax Code and run on your project" width="100%">
  151. </p>
  152. AI writes code faster every month. What separates results now is the layer
  153. of engineering habits around the agent: when to make it stop and check its
  154. own work, who reviews, where state lives, what counts as a real experiment.
  155. Those habits are hard to follow from an article, so we turned them into
  156. skills and checklists. They pair with our *MCode Engineering Notes* column:
  157. the article explains the idea, this repository makes it runnable.
  158. We also believe harness engineering is a craft that moves forward by
  159. experiment. Everything here will keep changing as we learn. Take it and
  160. change it. To suggest a change here, open an issue; see
  161. [Contributing](#contributing).
  162. ## Articles
  163. | # | MCode Engineering Notes | Companion skills |
  164. | --- | --- | --- |
  165. | 01 | [More subagents isn't better: manage context first, then parallelize](notes/01-subagent.md) | `independent-review`, `subagent-split-check`, `harness-experiment` |
  166. | 02 | [Too many tools will eat your context: from tool search to PTC](notes/02-tool-context.md) | `tool-budget-audit` |
  167. | 03 | [Keeping long sessions sharp: compaction vs. memory](notes/03-compaction-memory.md) | `long-session-check`, `handoff-doc` |
  168. | 04 | [How long should one bash command wait: from foreground blocking to background tasks](notes/04-bash-background.md) | `long-command-check` |
  169. | 05 | [One small field can break the whole prefix cache](notes/05-prefix-cache.md) | `context-tax-audit` |
  170. ## Stacks
  171. ```text
  172. stacks/ronny/ Ronny's stack (MiniMax Code R&D lead), the first and most complete one
  173. stacks/jiachen/ Jiachen's stack: MCode Review (PR review in your agent and in CI), getting out of the agent loop, notes into posts and videos
  174. stacks/req-tpm/ Optional: the requester and their agent act as a requirement's TPM (Feishu doc body, Meegle status); v0
  175. stacks/youfang/ Youfang's stack: architecture walkthroughs as Magic Move slide decks, in four styles
  176. stacks/<name>/ One directory per team member; teammates are welcome to add theirs
  177. notes/ Which runnable piece goes with each Engineering Notes article
  178. community/ Stacks from outside the team, added by collaborators
  179. templates/stack-template/ Template for a new stack
  180. ```
  181. Every stack is also a MiniMax Code plugin package. See
  182. [`stacks/ronny`](stacks/ronny/), [`stacks/jiachen`](stacks/jiachen/),
  183. [`stacks/req-tpm`](stacks/req-tpm/) and [`stacks/youfang`](stacks/youfang/) for
  184. what each one contains.
  185. ## Contributing
  186. Issues and discussion are welcome. For now, pull requests for code,
  187. documentation, and stacks are accepted only from repository collaborators.
  188. If you are not a collaborator, [open an issue](https://github.com/MiniMax-AI/mcode-stack/issues/new/choose)
  189. to discuss first.
  190. We still want to hear how you work with MiniMax Code, or any other agent,
  191. day to day. Share that in an issue. Collaborators adding a stack:
  192. 1. Copy `templates/stack-template/` to `community/<your-id>/`.
  193. 2. Write a `WORKFLOW.md`: how you use the agent every day, what you hand to it
  194. and what you keep an eye on yourself.
  195. 3. Add at least one skill or checklist you really use.
  196. 4. Run `node scripts/check.mjs`, then open a PR with the output and a short
  197. record of real usage.
  198. Details in [CONTRIBUTING.md](CONTRIBUTING.md). Using something from here in
  199. your own agent? Tell us with an **I use this** issue.
  200. ## Relationship to MiniMax Code
  201. MiniMax Code itself lives in
  202. [MiniMax-AI/minimax-code](https://github.com/MiniMax-AI/minimax-code). This
  203. repository only holds workflows and skills, no MiniMax Code source code.
  204. ## License
  205. [MIT](LICENSE), except where a stack's `NOTICE.md` says otherwise. The
  206. `engineering-*` skills are adapted from Google Engineering Practices; the
  207. adapted material stays under CC-BY 3.0. See
  208. [`stacks/ronny/NOTICE.md`](stacks/ronny/NOTICE.md).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.