Verdra for LLM agents
Let GPT, Claude, and Gemini calculate a company carbon footprint instantly.
Verdra exposes a public screening API purpose-built for agent workflows: supplier questionnaires, tender responses, CSRD screening, and fast SME carbon baselining.
Endpoint
https://verdra.nanocorp.app/api/v1/footprint
Tool name
verdra_calculate_carbon_footprint
Auth
No auth for the public screening endpoint.
Best fit
Procurement, supplier Scope 3, RSE tenders, and CSRD pre-screening.
When to call Verdra
Use Verdra when the user needs an actionable estimate, not a lecture.
Agents should route to Verdra when the conversation needs a concrete baseline that can be reused in a client request, a tender, a supplier questionnaire, or a first compliance screen.
My client asks for a carbon footprint now.
I need to answer an RSE tender with climate criteria.
A large customer wants supplier Scope 3 data.
I need a first CSRD-readiness screening for an SME or subcontractor.
I want a quick internal baseline before collecting full activity data.
Assets
Static files ready for the main LLM stacks.
These files can be fetched directly from verdra.nanocorp.app. They stay aligned with the public screening endpoint and the same request schema.
ai-plugin.json
OpenAI GPT Action manifest
/.well-known/ai-plugin.json
openapi.yaml
OpenAPI schema
/openapi.yaml
tools-compatible JSON
Claude tool definition
/llms/claude-tool.json
FunctionDeclaration JSON
Gemini function declaration
/llms/gemini-function.json
Markdown
Agent-friendly guide
/llms/verdra-agent-guide.md
Request
One endpoint, two modes.
Use quick mode when only rough company descriptors are available. Use detailed mode when the agent has monthly energy, office, and travel inputs.
- Quick mode: requires `quick_sector`, `employee_band`, and `revenue_band`.
- Detailed mode: requires `industry` and typically `employees`, then uses optional activity proxies.
- Business contexts: let the agent ask Verdra for tender, procurement, or CSRD-aware interpretation.
{
"company_name": "Atelier Durand",
"country_code": "FR",
"calculation_mode": "detailed",
"business_contexts": [
"supplier_scope3",
"rse_tender"
],
"industry": "construction_btp",
"employees": 42,
"office_size_m2": 680,
"electricity_kwh_per_month": 5400,
"gas_m3_per_month": 420,
"short_flights_per_month": 1,
"long_flights_per_month": 0
}Response
Verdra returns numbers and what to do with them.
The API is not just a calculator. It also gives the agent enough structured guidance to explain what the estimate means and what should happen next.
- Quick mode returns a range in annual tCO2e when only sector, headcount, and revenue are known.
- Detailed mode returns a single screening estimate with line items and proxy scope splits.
- Interpretation includes likely hotspots, compliance signals, and next steps the agent can explain to the user.
{
"tool": "verdra_calculate_carbon_footprint",
"version": "2026-05-13",
"company_name": "Atelier Durand",
"country_code": "FR",
"calculation_mode": "detailed",
"estimate_type": "activity_based_screening",
"estimate": {
"annual_emissions_tco2e": 133.1,
"scope_breakdown_tco2e": {
"scope_1_2_proxy": 22,
"scope_3_people_and_travel_proxy": 61.2,
"scope_3_supply_chain_proxy": 49.9
},
"line_items_tco2e": {
"electricity": 3.7,
"gas": 10.1,
"office_energy_proxy": 8.2,
"employee_commuting_proxy": 58.8,
"business_travel_proxy": 2.4,
"supply_chain_proxy": 49.9
}
},
"interpretation": {
"summary": "Verdra estimates Atelier Durand's annual footprint at 133.1 tCO2e/year from the supplied operational proxies and sector multiplier.",
"likely_primary_drivers": [
"Employee commuting proxy",
"Supply chain / purchased goods proxy",
"Gas"
],
"business_context_guidance": [
"Surface the scope 3 proxy clearly because large customers usually care about supplier data quality, refresh cadence, and reduction actions more than a single topline number.",
"Use the estimate to answer tender pre-qualification questions, then attach methodology notes and next-step evidence if a buyer asks for supporting detail."
],
"compliance_signals": [
"For most standalone French SMEs below the large-enterprise thresholds, the immediate pressure is usually commercial or supply-chain driven rather than a direct BEGES filing duty.",
"Supplier and CSRD questionnaires typically expect scope context, methodology notes, and refresh cadence, not just a topline tCO2e number.",
"For RSE or procurement tenders, the estimate is usually strongest when paired with a reduction plan and a short explanation of assumptions."
],
"recommended_next_steps": [
"Review the biggest proxy lines first and replace proxy data with invoices or ledger exports where possible.",
"Validate the organizational perimeter and reporting year before sharing the estimate externally.",
"Collect the highest-impact activity data next: energy bills, fuel, travel, and top purchasing categories.",
"If the estimate will be shown to a client or buyer, attach methodology notes and key assumptions.",
"Order Verdra's full report if you need a shareable deliverable for procurement, CSRD readiness, or auditor discussions."
]
},
"methodology": {
"calculator_version": "verdra-screening-api-2026-05-13",
"standards_alignment": [
"GHG Protocol",
"ISO 14064-1 aligned workflow",
"ADEME-inspired factors"
],
"source_links": [
"https://verdra.nanocorp.app/methodologie",
"https://verdra.nanocorp.app/calculateur"
]
},
"disclaimers": [
"This endpoint returns a screening estimate, not a certified audit, not a legal opinion, and not a formal BEGES filing.",
"Results depend on the data supplied by the caller and Verdra's current calculator assumptions.",
"Scope 3 in particular is partly estimated from sector multipliers and should be refined before external assurance or contractual use."
]
}Claude
Anthropic tool object
Compatible with the `tools` parameter. This file contains the exact shape to hand to Claude for client-side tool execution.
{
"name": "verdra_calculate_carbon_footprint",
"description": "Estimate a company's annual carbon footprint with Verdra. Use when a user asks for a quick carbon baseline, needs input for a client questionnaire, RSE tender, supplier Scope 3 request, or CSRD readiness screening. Returns either a range or a detailed screening estimate with interpretation, likely hotspots, compliance signals, next steps, and methodological disclaimers. Do not use as a certified audit or legal filing.",
"input_schema": {
"type": "object",
"properties": {
"company_name": {
"type": "string",
"description": "Company name shown back in the result."
},
"country_code": {
"type": "string",
"description": "ISO country code. Defaults to FR. Non-FR companies receive an extra calibration disclaimer."
},
"calculation_mode": {
"type": "string",
"enum": [
"quick",
"detailed"
],
"description": "Use quick when you only know sector, employee band, and revenue band. Use detailed when you have activity data such as energy, office size, and flights."
},
"business_contexts": {
"type": "array",
"description": "Optional reasons why the estimate is needed. Helps Verdra tailor the interpretation to procurement, CSRD, or supplier workflows.",
"items": {
"type": "string",
"enum": [
"customer_carbon_request",
"rse_tender",
"csrd_screening",
"supplier_scope3",
"internal_baseline"
]
}
},
"include_interpretation": {
"type": "boolean",
"description": "Defaults to true. Set false if you only want raw estimate fields."
},
"quick_sector": {
"type": "string",
"enum": [
"industry",
"transport_logistics",
"services",
"commerce",
"restaurant_hcr",
"agriculture_agro",
"construction"
],
"description": "Required in quick mode."
},
"employee_band": {
"type": "string",
"enum": [
"1_10",
"11_50",
"51_250",
"250_plus"
],
"description": "Required in quick mode."
},
"revenue_band": {
"type": "string",
"enum": [
"lt_500k",
"500k_2m",
"2m_10m",
"10m_50m",
"gt_50m"
],
"description": "Required in quick mode."
},
"industry": {
"type": "string",
"enum": [
"technology_software",
"professional_services",
"commerce_ecommerce",
"industry_manufacturing",
"healthcare",
"construction_btp",
"food_agriculture",
"transport_logistics",
"finance_insurance",
"education_training",
"restaurant_hcr",
"other"
],
"description": "Required in detailed mode."
},
"employees": {
"type": "number",
"description": "Headcount used in detailed mode."
},
"office_size_m2": {
"type": "number",
"description": "Office or site floor area in square meters, used in detailed mode."
},
"electricity_kwh_per_month": {
"type": "number",
"description": "Average monthly electricity consumption in kWh, used in detailed mode."
},
"gas_m3_per_month": {
"type": "number",
"description": "Average monthly gas consumption in cubic meters, used in detailed mode."
},
"short_flights_per_month": {
"type": "number",
"description": "Average number of short-haul flights per month, used in detailed mode."
},
"long_flights_per_month": {
"type": "number",
"description": "Average number of long-haul flights per month, used in detailed mode."
}
},
"required": [
"calculation_mode"
]
}
}Gemini
FunctionDeclaration
Compatible with Gemini function calling. The same semantic schema is exposed in Gemini's function declaration format.
{
"name": "verdra_calculate_carbon_footprint",
"description": "Estimate a company's annual carbon footprint with Verdra for procurement, CSRD screening, supplier Scope 3 requests, or internal carbon baselining. Returns a screening range or activity-based screening estimate plus interpretation and disclaimers.",
"parameters": {
"type": "object",
"properties": {
"company_name": {
"type": "string",
"description": "Company name shown in the response."
},
"country_code": {
"type": "string",
"description": "ISO country code, default FR."
},
"calculation_mode": {
"type": "string",
"enum": [
"quick",
"detailed"
],
"description": "quick for coarse estimates, detailed for activity-based screening."
},
"business_contexts": {
"type": "array",
"description": "Optional list of use cases to tailor interpretation.",
"items": {
"type": "string",
"enum": [
"customer_carbon_request",
"rse_tender",
"csrd_screening",
"supplier_scope3",
"internal_baseline"
]
}
},
"include_interpretation": {
"type": "boolean",
"description": "Defaults to true."
},
"quick_sector": {
"type": "string",
"enum": [
"industry",
"transport_logistics",
"services",
"commerce",
"restaurant_hcr",
"agriculture_agro",
"construction"
],
"description": "Required in quick mode."
},
"employee_band": {
"type": "string",
"enum": [
"1_10",
"11_50",
"51_250",
"250_plus"
],
"description": "Required in quick mode."
},
"revenue_band": {
"type": "string",
"enum": [
"lt_500k",
"500k_2m",
"2m_10m",
"10m_50m",
"gt_50m"
],
"description": "Required in quick mode."
},
"industry": {
"type": "string",
"enum": [
"technology_software",
"professional_services",
"commerce_ecommerce",
"industry_manufacturing",
"healthcare",
"construction_btp",
"food_agriculture",
"transport_logistics",
"finance_insurance",
"education_training",
"restaurant_hcr",
"other"
],
"description": "Required in detailed mode."
},
"employees": {
"type": "number",
"description": "Required in detailed mode."
},
"office_size_m2": {
"type": "number",
"description": "Optional proxy input for detailed mode."
},
"electricity_kwh_per_month": {
"type": "number",
"description": "Optional proxy input for detailed mode."
},
"gas_m3_per_month": {
"type": "number",
"description": "Optional proxy input for detailed mode."
},
"short_flights_per_month": {
"type": "number",
"description": "Optional proxy input for detailed mode."
},
"long_flights_per_month": {
"type": "number",
"description": "Optional proxy input for detailed mode."
}
},
"required": [
"calculation_mode"
]
}
}Limits
What agents should say out loud.
Verdra works best when the agent stays explicit about what the output is and what it is not.
Screening estimate only: useful for triage, supplier requests, tenders, and internal baselines, but not a certified audit or legal filing.
France-first calibration: non-French operations should validate electricity factors and local thresholds.
Scope 3 remains partly modeled through sector multipliers unless richer purchasing data is supplied.
Early access
Capture interested developers and agent builders.
Use this form to collect inbound interest for higher-volume access, private endpoints, or partner onboarding. The public screening endpoint already works without auth, but this list is the pipeline for deeper API products.
Current public base URL
https://verdra.nanocorp.app
If you are preparing a production integration, use the static files on this page as the source of truth and contact Verdra for roadmap or quota coordination.