Which accounts are missing QBR preparation evidence?

Inspect the worked example

Our synthetic 130-account fixture selects 120 scheduled reviews: 90 have all required evidence categories and 30 have gaps. 10 have stale usage evidence, 10 future-only observations and 10 missing usage evidence. 10 later reviews are excluded.

Read the free six-account report excerpt · See package contents and SHA256

These are invented accounts and documents, not customer results.

One reusable source kit

19 USDC on Base · fixed edition 4 October 2026

Includes original Python source, nine regression checks, guide, MIT licence, synthetic roster/evidence CSVs and the complete expected 130-account report. MIT permits modification and redistribution. No exclusivity or recurring support is included.

Paid ZIP endpoint · Client discovery

Use an authorized funded x402 client to GET /store/qbr-coverage.zip. An ordinary browser link returns payment requirements; it does not sign or buy. Live paid settlement and native Dot execution remain unverified.

python3 -B qbr_coverage.py --roster examples/roster-130.csv --evidence examples/evidence-130.csv --as-of 2026-10-04 --categories usage,outcomes,open_actions --output report-130.json

Python 3.10+; no installation or API key. An agent needs Python and authorized local file access. The tool is not a CRM integration or a hosted review service.

Explicit rules, bounded inputs

Two UTF-8 CSVs: roster account_id,review_date and evidence evidence_id,account_id,category,observed_date,reference. Supply an as-of date and 1–12 required categories. Default 90-day inclusive evidence window and 30-day review horizon; both configurable.

Up to 1000 accounts, 10,000 evidence records and 2 MB per file. Exact headers, unique IDs and real calendar dates. Orphan and unrequested records are flagged. Future observations never fill a gap. Output is JSON; existing output paths are protected from overwrite.

The kit never opens a source reference, contacts customers, uploads files, scores account health or verifies QBR completion. Complete supplied coverage does not mean a correct or complete QBR. Missing evidence can mean an incomplete export. Follow your own customer-data access rules and keep reports in the authorized environment.

Free alternative: compare the roster and evidence inventory manually with a spreadsheet or your own script. This purchase packages the original implementation and reproducible tests. No OpenAI or CRM vendor affiliation; no paid buyers or endorsements claimed.