Skip to main content
Back to Prompt Library
implementation

Build the application

Inspect the original prompt language first, then copy or adapt it once you know how it fits your workflow.

Linked challenge: Build an Open Public Procurement Catalog Verification Agent

Format
Code-aware
Lines
94
Sections
11
Linked challenge
Build an Open Public Procurement Catalog Verification Agent

Prompt source

Original prompt text with formatting preserved for inspection.

94 lines
11 sections
No variables
1 code block
## Build
Screen supplier product submissions for public procurement bids against mandatory technical specifications. Compile verifiable compliance matrices and audit trails for procurement evaluators.

## User
Public Procurement Technical Evaluator

## Workflow
1. Ingest supplier specification sheets and test certificate records for bid items.
2. Extract standardized engineering properties and environmental compliance markers.
3. Compare extracted item specifications against published public tender clauses.
4. Assemble structured evaluation scorecards with verifiable line-item citations.

## Interface
Bid review dashboard highlighting specification deviations, certificate expiry dates, and clause mappings.

## Data
Curate mock tender specifications and synthetic vendor bid packets in normalized JSON formats.
Starter examples below are synthetic. They are not downloaded records.

## Source
[ONDC network extension](https://raw.githubusercontent.com/ONDC-Official/protocol-network-extension/main/README.md)
Applies ONDC network extension contracts and category taxonomy rules to public procurement items.

## Tools
crewai, pydantic, tabulate.
Orchestrates dedicated evaluation agents for technical specs, certification validity, and clause compliance.
Local or hosted LLM with high accuracy in structured numerical and standard comparison.

## Deliverables
Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

## Acceptance
- Matches supplier parameters against minimum tender thresholds with exact delta tracking.
- Identifies missing mandatory testing documentation.
- Generates audit-ready evaluation summaries ready for procurement officer sign-off.
- Rejects bids containing unverified specification claims.

## Starter Examples
```json
[
  {
    "input": {
      "tender_id": "TND-2026-MED-04",
      "clause_requirements": {
        "min_battery_hours": 8,
        "certifications": [
          "ISO-13485"
        ],
        "warranty_months": 24
      },
      "bidder_submission": {
        "vendor_id": "VEND-881",
        "battery_hours": 10,
        "certifications_attached": [
          "ISO-13485",
          "CE"
        ],
        "warranty_months": 36
      }
    },
    "expected_output": {
      "compliance_result": "COMPLIANT",
      "passed_clauses": 3,
      "failed_clauses": 0,
      "audit_verdict": "ELIGIBLE_FOR_COMMERCIAL_OPENING"
    }
  },
  {
    "input": {
      "tender_id": "TND-2026-MED-04",
      "clause_requirements": {
        "min_battery_hours": 8,
        "certifications": [
          "ISO-13485"
        ],
        "warranty_months": 24
      },
      "bidder_submission": {
        "vendor_id": "VEND-882",
        "battery_hours": 6,
        "certifications_attached": [
          "CE"
        ],
        "warranty_months": 12
      }
    },
    "expected_output": {
      "compliance_result": "NON_COMPLIANT",
      "passed_clauses": 0,
      "failed_clauses": 3,
      "disqualifications": [
        "BATTERY_BELOW_MINIMUM",
        "MISSING_ISO_13485",
        "WARRANTY_BELOW_MINIMUM"
      ],
      "audit_verdict": "REJECTED_TECHNICAL_STAGE"
    }
  }
]
```

## Safety
Handles synthetic tender fixtures only; generates recommendations without issuing binding government contract awards.
Use mock external writes. Require human review before any real deployment.

Adaptation plan

Keep the source stable, then change the prompt in a predictable order so the next run is easier to evaluate.

Keep stable

Hold the task contract and output shape stable so generated implementations remain comparable.

Tune next

Update libraries, interfaces, and environment assumptions to match the stack you actually run.

Verify after

Test failure handling, edge cases, and any code paths that depend on hidden context or secrets.