Requirement Intake (req-intake)
Challenges whether a new requirement is worth doing, forces a verifiable expected outcome, then creates a Feishu requirement doc and a Meegle work item cross-linked to each other.
What it does
- Challenge first, record second: walks the requester through who hits what problem in what situation (real evidence required), what happens if you don't build it, whether a smaller fix exists, and who is involved. Only three verdicts are allowed: proceed / recommend not doing / gather more info.
- Forces an Expected Outcome table: metric or observable behavior, baseline with its source, target, how and who measures it, observation window. Vague phrasing like "better experience" is rejected.
- Creates and cross-links artifacts: builds the Feishu requirement doc from a template via
lark-cli, creates the Meegle (Feishu Project) work item via themeegleCLI, then writes each link back into the other. - Acceptance checklist: the work item reads back the doc link, the doc shows the Meegle link, and every outcome row has a baseline (or measurement plan) plus observation window.
Who it's for
- TPMs, PMs and tech leads who run product requirements on Feishu Docs + Meegle
- Teams whose one-line chat requests vanish because no doc and no ticket ever gets created
- Orgs that want success criteria quantified at intake, not at release
Prerequisites:
lark-cliand themeegleCLI must be installed and authenticated; the skill relies ontemplates/requirement-doc.mdandtemplates/meegle-field-mapping.md.
Examples
- A teammate posts "the
/resumelist is slow, can we add caching?" → the skill demands real evidence, fills the table with baseline 2.4s (last week's telemetry) → target ≤ 800ms, 7-day post-release window, then creates the doc and work item. - The request is actually solvable by a config change → step 3 flags it as a smaller fix and closes with a documented "recommend not doing" verdict.
- A doc exists but no ticket → creates the Meegle work item and replaces "Work item: to be created" in the doc with the real link.
· · · Install guide · · ·
Try it now, no install
Paste this into Claude to use the skill without installing anything.
Read the instructions in this file and follow them to help me: https://raw.githubusercontent.com/MiniMax-AI/mcode-stack/HEAD/stacks/req-tpm/skills/req-intake/SKILL.md What I want: (describe your task here)
If Claude can't open the link, open it yourself and paste the contents instead.
↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on AI Skill Mart. Copy the stacks/req-tpm/skills/req-intake folder from the GitHub repo MiniMax-AI/mcode-stack into my ~/.claude/skills/req-intake/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/MiniMax-AI/mcode-stack.git && mkdir -p ~/.claude/skills && cp -r mcode-stack/stacks/req-tpm/skills/req-intake ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal and clone the repo:
git clone https://github.com/MiniMax-AI/mcode-stack.git - Create the skills folder:
mkdir -p ~/.claude/skills - Copy the skill in:
cp -r mcode-stack/stacks/req-tpm/skills/req-intake ~/.claude/skills/ - Confirm
~/.claude/skills/req-intake/templates/containsrequirement-doc.mdandmeegle-field-mapping.md. - Install the dependencies: the
lark-clitool (Feishu docs) and themeegleCLI, then log in / authenticate both. - Restart Claude Code and say something like "intake this requirement" (or the Chinese triggers 提需求 / 我想要 / 能不能加) to activate the skill.
- Once per session, run
meegle project searchfor the project key andmeegle workitem meta-typesfor the work item type so the later steps run cleanly.