Before: An RFP requirement happens to describe one vendor's product.
Requirements stay outcome-based
Brand-specific language must have an approved source or be rewritten around the business need.
HINT for Procurement
Your assistant drafts RFPs fast. HINT keeps them fair and source-based.
HINT keeps requirements, timelines, evaluation rules and approvals in plain-text notes beside tender documents. Your assistant reads the rules for this procurement before drafting supplier-facing language.
The package is named for the role: @openhint/hintbook-procurement-specialist.
Where it helps
Three situations your team will recognize
Before: Internal evaluation weights leak into the supplier document.
Scoring stays in its proper boundary
Internal evaluation context is available to the panel without being emitted into the RFP.
Before: A budget figure appears because the assistant assumed a range.
Commercial facts require authority
Missing budgets and deadlines become questions for the procurement owner rather than invented commitments.
See it
A real evaluation requirement from the demo
The Springfield procurement demo grounds an evaluation requirement in SRC-1.
demo-springfield-procurement/tenders/monorail/evaluation.md.hint ↗
# requirement overview-1 {#procurement_specialist_overview}
Approved requirement grounded in SRC-1.
# source SRC-1
../../sources/approved.mdWhat your assistant does next
It writes neutral evaluation language grounded in the approved source and keeps internal scoring details out of supplier-facing material.
What your AI assistant receives
<requirement name="overview-1">
Approved requirement grounded in SRC-1.
</requirement>
<source name="SRC-1">../../sources/approved.md</source>Your boundaries stay visible
What it will never do
- Invent a budget, deadline, approval or vendor fact.
- Write a brand-specific requirement without a justified source.
- Expose confidential evaluation weights in the RFP.
FAQ
Questions procurement specialists ask
Does it select a supplier?
No. It keeps criteria and evidence structured; authorized evaluators make and approve the decision.
How is vendor neutrality enforced?
Declare outcome-based requirements and prohibit unsupported brand references; reviewers still assess legitimate exceptions.
Do weights have to total 100?
Only if your declared evaluation policy says so. HINT does not invent a scoring rule.
For your technical colleague
The vocabulary
A hintbook is a vocabulary for your profession—installed, not written by you.
tender- Sets the procurement, authority and document boundary.
need- States the business outcome behind the purchase.
budget- Records an authorized budget from a named source.
requirement- States a supplier-facing need with a stable identifier.
exclusion- Marks an internal rule that must not enter the RFP.
question- Records an approved supplier clarification.
criterion- Names a factor evaluators apply consistently.
weight- Records confidential scoring importance where authorized.
Manual setup
Bootstrap is read-only: it prints instructions for the assistant. The assistant performs the installation.
npm install -g @openhint/cli
hint config
hint add @openhint/hintbook-procurement-specialist
hint apply
hint verify tenders/monorail/evaluation.md
hint emit tenders/monorail/evaluation.md