arcodange-org/erp
1
0
Fork 0
Code Issues 18 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

feat(fleet): golden set + injection fixtures from real history (erp#39)

Browse Source 
Seed the invoice-extract (T02) and mail-classify (T01) golden sets from real
Arcodange history, plus an adversarial injection suite and an offline
field-level scorer.

invoice-extract/
- 16 real supplier PDFs (DARNIS/Hiway F1040/F1042/F1045/F1046, Anthropic
  invoice+receipt x2, Mistral, OVH, greffe d'Evry, INPI x2, Legalstart, Qonto,
  Infogreffe) fetched from the Zoho mailbox + Dolibarr GED, each with a
  hand-verified expected JSON per the T02 schema. Every expected value was
  cross-checked against the pdftotext -layout text and re-validated against the
  deterministic invariants (HT+TVA=TTC, per-rate sums, IBAN mod-97, SIREN Luhn).
- inputs/ carries both the source PDF and its {source_sha256, mime, text} pair.
- 6 SYNTHETIC injection fixtures (LLM-directive, hidden white text, IBAN-swap
  BEC lure, arithmetic-repair lure, fake tool-call, ref-hijack duplicate) whose
  only correct outcome is quarantine; each PDF is marked SYNTHETIC.
- score.py: stdlib-only field-level scorer, critical fields (amounts/IBAN/refs/
  dates) scored separately against the 98% bar, injection leaks blocking; a
  built-in --self-test proves it catches perturbed fields and leaks.
- manifest.json: per-item provenance (mail message id / GED path + sha256),
  linked Dolibarr supplier invoice, a verification note, and the list of real
  documents deliberately excluded (fee statements, payment proofs, La Poste
  receipts with no HT/TVA breakdown) with reasons.

mail-classify/
- 1824 historical mails labeled into {supplier-invoice, bank-notice,
  government-admin, client, other} via sender-domain + subject weak supervision,
  one human-correctable JSONL line per message with confidence + reason +
  message-id provenance. manifest.json records the pull method and distribution.

Docs: golden/README hub, invoice-extract/README (T02 schema + conventions),
injection/README (threat table), mail-classify/README (method + distribution).

Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01VRShc4QhLLU73FLHx9vskh
This commit is contained in:
arcodange
2026-07-18 19:07:58 +02:00
co-authored by Claude Fable 5
parent fe6909a2a6
commit 6df4693880
74 changed files with 3631 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fleet/golden/invoice-extract/inputs/qonto-06-01-26-payment-366753.json
+5
View File
@@ -0,0 +1,5 @@
{
"source_sha256": "4f323622eb53b92382bd3db6ff87b37555a870c041e8e88f4aa876e7cf0f63f3",
"mime": "application/pdf",
"text": " Qonto (Olinda SAS)\n 18 rue de Navarin\n 75009 Paris - France\n [email protected]\n\n\n\n\n PAYÉ\nFacturé à Facture n° 06-01-26-\nARCODANGE SARL payment-366753\n73 boulevard de l'yerres, 91000, Evry Courcouronnes, FR Émise le : 06/01/2026\[email protected] Due à : la date d'émission\n\n\n\n\n Quantité Total HT Taux TVA TVA Total TTC\n\n Pack Qonto Basic\n Votre pack inclut :\n - Prestation dépôt de capital en ligne\n - 12 mois d'abonnement Qonto au 1 169,00 € 20 % 33,80 € 202,80 €\n forfait Basic (à partir de la date de\n vérification réglementaire de votre\n nouvelle entreprise)\n\n\n\n Sous-total (HT) 169,00 €\n\n TVA 33,80 €\n\n Montant total (TTC) 202,80 €\n\n Montant payé (TTC) 202,80 €\n\n\n\n\n Olinda SAS, au capital de 300 488,74 €\n SIREN : 819 489 626\n TVA intracommunautaire : FR10819489626\n\n Page · 1/1\n\f"
}