B2G Skill — Multi-Tool Reference (Sell to Government)
Purpose
This skill provides the B2G revenue playbook: selling to national/local governments via tenders/licitaciones and public-private partnerships (PPP). It is country-agnostic — the pattern works across procurement platforms: SICOP (CR), CompraNet (MX), Guatecompras (GT), SECOP (CO), ComprasNet (BR), COMPR.AR (AR), Mercado Público (CL), and more. The reference implementation lives in scripts/b2g/.
Capabilities
- Tender normalization:
tender_normalizer.py— normalize raw records from any procurement platform →Tendermodel (source, number, description, institution, UNSPSC, published/deadline) - UNSPSC mapping:
unspsc_map.py— map the client's catalog (products/services) → UNSPSC 8-digit codes using an ES/EN synonym dictionary (same pattern as LicitaNext/hunter-core) - Match + scoring:
tender_match.py— score tenders by UNSPSC overlap and closing urgency - RFP analysis:
rfp_analyzer.py— extract actionable requirements (habilitante/documental/técnico/financiero/compliance) → agent-ready work - Procurement landscape: references to platforms per country, UNSPSC, RFP reading, PPP, compliance
Usage
This skill enables B2G guidance for agents across Claude Code, Cursor, and OpenCode. Given "find public tenders matching our catalog / analyze this RFP / structure a PPP offer", run the scripts/b2g/ pipeline and feed output to revenue-ops-triage for prioritization and pricing-strategist for offer strategy.
Pipeline
flowchart LR
PL[Procurement platform] --> N[tender_normalizer]
N --> T[Tender]
CAT[Client catalog] --> U[unspsc_map]
U --> M[tender_match]
T --> M
M --> R[rfp_analyzer]
R --> TRI[revenue-ops-triage]
References
references/procurement-platforms.md— platform matrix per countryreferences/unspsc.md— classification standard + ES/EN dictionary approachreferences/rfp-requirements.md— how to read a pliego/RFPreferences/public-private-partnership.md— PPP / alianzas público-privadasreferences/compliance.md— fianza, firma digital, certifications per country