Methodology v1.1 — Frozen
NOXAR Methodology
Evidence-first methodology for analyzing public construction tenders.
Principles
- Evidence first — every score requires a document citation
- Score justified — rubric rows only, no discretionary points
- No opinion without reference — unsourced claims are flagged
- Reproducibility ±3 points — two analysts on the same DCE should align
Tender Health Score (0–100)
Evidence-backed composite score. Recommendation follows the THS band unless a Hard Veto applies.
| Range | Recommendation |
|---|---|
| ≥ 80 | GO |
| 60–79 | GO WITH CONDITIONS |
| 40–59 | HIGH CAUTION |
| < 40 | NO GO |
Hard Veto Logic
Ten binary rules (HV-01 to HV-10). Any trigger → NO GO regardless of THS.
- •HV-04: RC, CCAP, or CCTP missing from acquired DCE
- •HV-06: Missing referenced contractual document (CCAP Art. 2/3)
- •HV-07: Schedule contradiction >8 weeks without explanation
- •HV-09: Submission deadline passed
All vetoes are audited before final recommendation. False positives require evidence review.
Confidence Labels
HIGH — ≥90% sub-criteria evidence-backed, 0 vetoes
MEDIUM — <20% at zero, ≤2 ambiguities, no veto
LOW — Vetoes, unread files, or ≥3 ambiguities
BLOCKED — Cannot compute THS
Evidence Layer
Each sub-criterion includes: score, factual DCE statement, and reference (filename — article — page).
Payment Terms Score: 3/5 (3/5 pts) Evidence: 50-day global payment period. Reference: RC/reglement_consultation.pdf — §1.8 — page 4