Master Document Numbering & Metadata Assistant: No More Gaps, Clashes, or Mystery IDs
Turn your numbering scheme and your live register into a tool that hands out the next correct document ID, captures the metadata every doc needs, and refuses to create a duplicate or out-of-pattern number — with the document controller approving before any ID is reserved.
A web tool where you define your numbering scheme once, then anyone requesting a new document gets the next correct ID proposed automatically with required metadata validated, the document controller reviews and approves, the ID is reserved in the register, and you export the full updated numbering ledger as a CSV.
Before you start
- A Supabase account (free)
- A Vercel account (free)
- A Resend account (free)
- Your numbering scheme / pattern written down
- A CSV or Google Sheet of the document numbers already in use
- Claude Code or any AI coding agent
The problem this kills
Document control should be the calm corner of your quality system. Instead it's a spreadsheet held together by one person's memory. Someone needs a number for a new SOP, so they ping the document controller, who scans a long register, eyeballs the pattern, picks what looks like the next free ID, and types it back — along with whatever metadata they remembered to ask for. Multiply that by every department and every site, and you get the predictable mess: two documents with the same number, gaps where a number was promised but never used, IDs that break the pattern (a stray SOP-QA-07a in a scheme that only allows three digits), and a register where half the rows are missing an owner, a classification, or a retention period.
When the auditor pulls the thread, you're explaining why WI-MFG-112 exists twice and why nobody knows who owns POL-HR-005. You don't need to keep hand-assigning numbers and hoping, and you don't need to be a developer to fix it.
What you'll build
A simple internal web tool. First you teach it your numbering scheme — the pattern, the parts (document type, department, site), the number width, and any retired or reserved ranges. You import the document numbers already in use from your CSV or Sheet so it knows reality. Then, when someone requests a new document, the tool proposes the next correct ID for that type/department/site, checks it against the whole register so it can't collide or break the pattern, and forces the required metadata — owner, classification, retention, related process — to be filled before it will go any further.
The document controller sees the proposed number and metadata on a clean review screen, fixes anything off, and clicks Approve. Only then is the ID reserved in the register so no one else can grab it. At any point you can export the full numbering ledger as a CSV in exactly the columns your register expects.
What's inside the Implementation Plan
The downloadable plan is a step-by-step file you paste into an AI coding agent. It opens by interviewing you about your business — what your numbering pattern actually looks like, the exact type/department/site codes you use, how wide the running number is, which ranges are retired or reserved, the precise metadata fields you require and their allowed values, your typical and peak request volumes, and the messy edge cases that always trip you up — and then it tailors the data model, the validations, and every later step to your answers. This is not a generic template; the agent reflects a short spec back to you and waits for your thumbs-up before it builds anything.
From there it walks the agent through importing your existing register, encoding your pattern as a validator, building the request-and-propose flow, the metadata required-field checks, the controller review-and-approve screen, the ID reservation, and the full-ledger CSV export — each step with a ready-to-copy prompt. There's also a fallback so you can build and run the whole thing today even with no API to your existing document management system.
The governance it includes (this is the point)
This is real document-control tooling, so it ships with the controls a quality team needs: login so only your team can use it, row-level security so you only ever see your own organization's register, a complete audit trail of who requested, edited, approved, and reserved every number and when, a hard human-approval gate so no ID is written into the register until the document controller signs off, and duplicate guards keyed on the document number so the same ID can never be reserved twice or break the pattern. Retired and reserved ranges are honored automatically, so a number that's off-limits is never proposed.
Who it's for
Document controllers and quality admins who hand-assign numbers and end up with gaps, clashes, and inconsistent metadata. If you can describe how your numbering scheme works and what metadata each document needs, you can build this.
You've got this — start with the plan, paste the first prompt, answer the interview, and you'll watch your numbering chaos turn into a clean, guarded register the same afternoon.