Ghg Ledger

Official6 tools

by jdhart81 · Python

Deterministic Scope 1/2/3 GHG inventories with bundled factors and audit hashes.

Captured live from the server via tools/list.

calculate_inventory

Calculate an auditable GHG inventory from explicit activity records. Returns per-line gas/CO2e results, Scope 1/2/3 and category rollups, dual Scope 2 reporting, indeterminate lines, factor lineage, notary payload, and an offset-clearinghouse dry-run weave. This is calculation, not advice.

Parameters2

activities	array	required
options	any	optional

Raw schema

{
  "type": "object",
  "properties": {
    "activities": {
      "items": {
        "additionalProperties": true,
        "type": "object"
      },
      "title": "Activities",
      "type": "array"
    },
    "options": {
      "anyOf": [
        {
          "additionalProperties": true,
          "type": "object"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Options"
    }
  },
  "required": [
    "activities"
  ],
  "title": "calculate_inventoryArguments"
}

classify_activity

Suggest scope and Scope 3 category from the bundled deterministic map. No inference is used; unknown activity types return no suggestion.

Parameters1

activity

object

required

Raw schema

{
  "type": "object",
  "properties": {
    "activity": {
      "additionalProperties": true,
      "title": "Activity",
      "type": "object"
    }
  },
  "required": [
    "activity"
  ],
  "title": "classify_activityArguments"
}

list_factor_packs

List the bundled pack version/digest, supported regions and years, activity types, GWP set, and explicit MVP coverage.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "list_factor_packsArguments"
}

get_factor_pack

Return factors, GWP values, conversions, sources, and pack SHA for one exact bundled region/year. No nearest-region or nearest-year substitution.

Parameters2

region	string	required
year	integer	required

Raw schema

{
  "type": "object",
  "properties": {
    "region": {
      "title": "Region",
      "type": "string"
    },
    "year": {
      "title": "Year",
      "type": "integer"
    }
  },
  "required": [
    "region",
    "year"
  ],
  "title": "get_factor_packArguments"
}

verify_result

Recompute a prior inventory's audit hash and conservation checks. Pass the prior result object as JSON; tampering or stale factor lineage is flagged.

Parameters1

result_json

string

required

Raw schema

{
  "type": "object",
  "properties": {
    "result_json": {
      "title": "Result Json",
      "type": "string"
    }
  },
  "required": [
    "result_json"
  ],
  "title": "verify_resultArguments"
}

describe_agent

Return fleet-standard capabilities, version, pack digest, and pricing.

Parameters

No parameters.

Raw schema

{
  "type": "object",
  "properties": {},
  "title": "describe_agentArguments"
}

README.md26 KBView on GitHub

Viridis Agent Fleet — MCP Surface

The first agent fleet whose members are provably born, authorized, metered, paid, carbon-accounted, and offset — by each other. Read the receipts. Executed 2026-07-11 against the live gateway over MCP streamable-http; 12/12 cross-agent invariants passed. The rails it ran on are live at the same endpoint, free to call.

The public MCP-facing surface of the Viridis agent stable: a 19-agent agent-to-agent (A2A) economy published on the Model Context Protocol registry under the io.github.jdhart81 namespace. This repository is the callable spec + provenance for those listings — the registry manifests, the JSON-Schema tool definitions, the public agent contracts, and the reference gateway. The agent cores are maintained privately; everything here is what a calling agent needs.

By Viridis LLC — conservation technology.

The economy: identity → trust → escrow → settlement → constitution

Trustworthy agents transacting with agents needs rails. This fleet ships them as composable MCP services:

Layer	Agent	Endpoint	What it provides
Identity	`agent-identity-registry`	`/identity/mcp`	Content-addressed agent DIDs + capability discovery — the passport + directory
Trust	`agent-trust-oracle`	`/trust/mcp`	Decay-weighted reputation + tamper-evident trust attestations
Settlement	`agent-escrow`	`/escrow/mcp`	Trustless escrow with an exactly-once state machine + audit hash chain
Metering	`agent-metering`	`/metering/mcp`	Usage metering + SLA accounting — the meter behind x402
Arbitration	`agent-arbitration`	`/arbitration/mcp`	Dispute-resolution oracle consuming trust signals
Compute ledger	`agent-compute-ledger`	`/compute-ledger/mcp`	"Compute is carbon" accounting plus an immutable GHG-inventory chain
Covenant	`agent-covenant`	`/covenant/mcp`	Deny-by-default authority leases for agents wielding real power
Provenance	`agent-provenance`	`/provenance/mcp`	Genesis lineage plus a content-addressed artifact DAG
Offsets	`agent-offset-clearinghouse`	`/offsets/mcp`	Verified-credit carbon accountability — the conservation flywheel
Interop	`agent-erc8004-bridge`	`/erc8004/mcp`	MCP-native bridge to ERC-8004 identity and reputation
Surety	`agent-surety`	`/surety/mcp`	Deterministic bonding and ruling-gated slashing
Notary	`agent-notary`	`/notary/mcp`	Commit-reveal delivery proofs and verification
Discovery	`wavefunction-search`	`/wavefunction/mcp`	Demand-side agent and collective discovery
Revenue	`smartscale`	`/smartscale/mcp`	Credit-card-calibrated measurement — the first sellable service
Revenue	`protogen`	`/protogen/mcp`	MCP CAD services; bundles with SmartScale (measure → CAD)
Revenue	`regulatory-radar`	`/regulatory-radar/mcp`	CSRD/TNFD compliance-as-a-service
Revenue	`taxcredit-engine`	`/taxcredit-engine/mcp`	Auditable 45Q/45V/45Y/48E/45X scenarios
Revenue · climate	`ghg-ledger`	`/ghg-ledger/mcp`	Deterministic Scope 1/2/3 inventories with dual Scope 2 and audit hashes
Enabler	`narrative-engine`	`/narrative-engine/mcp`	Grant / investor / policy narrative generation

Federated member: EnergyAI — energy intelligence (solar estimates, US incentives by ZIP, Energy Node Scores) on its own infrastructure, discoverable through this fleet's catalog.

Calling an agent

Every agent is a streamable-HTTP MCP server at https://mcp.viridisconservation.com/<path>/mcp. Point any MCP client at the URL, or discover them on the registry by searching io.github.jdhart81. Machine-readable fleet catalog: /.well-known/ai-catalog.json · Fleet health: /healthz. Each tools.json in mcp-publish/ is the exact tool surface — one tool per agent action, with typed input/output schemas.

Tools

The aggregate bridge exposes 131 namespaced tools. Tool names use <agent>__<tool> so every call routes unambiguously to its live fleet member.

Tool	Description
`identity__register_agent`	Register (or idempotently update) an agent identity. Returns a deterministic content-addressed DID. capabilities is a non-empty list of lowercase capability tags other agents can discover you by.
`identity__resolve_agent`	Resolve an identity by agent_id or DID to its full public registration.
`identity__discover_agents`	Find ACTIVE agents matching ALL requested capabilities (AND semantics), deterministically ordered by match count then reputation.
`identity__revoke_agent`	Revoke an identity: it disappears from discovery (terminal) but its record is retained for auditability.
`identity__list_registrations`	List registrations, optionally filtered by status (ACTIVE\|REVOKED).
`identity__describe_agent`	Fleet-standard self-description.
`trust__record_outcome`	Record an interaction outcome for an agent. kind: success \| delivered \| dispute_won \| failure \| undelivered \| dispute_lost \| timeout \| security_incident (security incidents carry a 3x penalty).
`trust__score_agent`	Get an agent's decay-weighted trust score in [0,1] and tier. Unknown agents get a neutral 0.5 prior — no blind trust, no unfair zero.
`trust__attest`	Issue a tamper-evident (hash-chained) trust attestation for an agent.
`trust__verify_attestation`	Verify a previously issued attestation by recomputing its hash.
`trust__history`	Full outcome history + attestation count + current score for an agent.
`trust__describe_agent`	Fleet-standard self-description.
`escrow__open_escrow`	Open an escrow between payer and payee. amount_minor is a positive integer in minor units (cents). The platform fee is computed and FROZEN at open (ceil bps). Returns the escrow_id (state OPEN).
`escrow__fund_escrow`	Mark an OPEN escrow as FUNDED (idempotent). payment_ref links the payment-rail transaction.
`escrow__release_escrow`	Release a FUNDED/DISPUTED escrow to the payee (exactly-once — a repeat release returns the existing terminal record, never a double payout).
`escrow__refund_escrow`	Refund an OPEN/FUNDED/DISPUTED escrow to the payer (exactly-once). Refunding an OPEN escrow is a cancel.
`escrow__dispute_escrow`	Move a FUNDED escrow to DISPUTED. An arbiter (agent-arbitration-agent) then resolves it to release or refund.
`escrow__escrow_status`	Current record for an escrow.
`escrow__list_escrows`	List escrows, optionally filtered by state (OPEN\|FUNDED\|RELEASED\|REFUNDED\|DISPUTED).
`escrow__verify_audit`	Validate the tamper-evident audit hash chain for an escrow.
`escrow__describe_agent`	Fleet-standard self-description.
`metering__create_meter`	Create a usage meter between a provider and a consumer agent. unit: what is being counted (call, token, kwh, ...). Price is in minor currency units (cents) per unit. Returns the meter_id.
`metering__record_usage`	Record a usage event. Idempotent on event_id (safe to retry — never double-billed). outcome is 'ok' or 'error' and feeds the SLA report.
`metering__usage_summary`	Totals for a meter: event count, total quantity, accrued minor units.
`metering__sla_report`	Pure SLA report: success_rate vs sla_target, breach flag. No mutation.
`metering__close_period`	Freeze all open events into an immutable invoice (exactly-once). The invoice amount is what agent-escrow-agent should settle.
`metering__verify_chain`	Verify the tamper-evident event hash chain for a meter.
`metering__list_meters`	List all meters with event/invoice counts.
`metering__describe_agent`	Fleet-standard self-description.
`arbitration__file_case`	File a dispute over an escrow. Opens the evidence window. Parties must be distinct; amount is in minor currency units (cents).
`arbitration__submit_evidence`	Submit evidence while the case is open. kind: delivery_proof (weight 3), log (2), or statement (1). Only the named parties may submit.
`arbitration__set_trust_scores`	Attach trust-oracle reputation scores (party -> [0,1]) as ruling inputs.
`arbitration__rule`	Issue the deterministic ruling: allocates 100% of the disputed amount from evidence weights + trust scores, and emits an escrow instruction (release/refund). Exactly-once — re-ruling returns the existing ruling.
`arbitration__verify_ruling`	Recompute the ruling from its cited evidence + trust inputs and check it matches the stored allocation (machine-checkable justice).
`arbitration__get_case`	Fetch the full case record, including any ruling.
`arbitration__list_cases`	List cases, optionally filtered by state (FILED\|EVIDENCE_OPEN\|RULED).
`arbitration__describe_agent`	Fleet-standard self-description.
`compute-ledger__record_work`	Record a unit of agent compute work. energy_j = power_w * duration_s; carbon_g follows from grid intensity. If bit_ops is declared, the entry is validated against the Landauer floor (bit_ops * kB * T * ln2) — physically impossible claims are rejected. Idempotent on entry_id.
`compute-ledger__footprint`	Aggregate footprint for an agent: total J, kWh, gCO2e, cost, and mean Landauer efficiency. Totals are exact sums of the ledger entries.
`compute-ledger__attest`	Issue a content-addressed attestation for a ledger entry (verifiable green-compute / energy claim).
`compute-ledger__verify_attestation`	Verify an attestation by recomputing the entry hash.
`compute-ledger__verify_chain`	Verify the tamper-evident hash chain of an agent's ledger.
`compute-ledger__list_entries`	List all ledger entries for an agent.
`compute-ledger__record_inventory`	Record an audited GHG inventory in a separate append-only chain. mass_g is exact integer grams; content and factor-pack digests are bare lowercase SHA-256 hex. Idempotent on inventory_id.
`compute-ledger__get_inventory`	Fetch one immutable inventory record by inventory_id.
`compute-ledger__list_inventories`	List the separate GHG inventory chain for an agent.
`compute-ledger__verify_inventory_chain`	Verify an agent's inventory hash chain independently of compute work.
`compute-ledger__describe_agent`	Fleet-standard self-description.
`smartscale__credit_card_photo_instructions`	Return user-facing capture instructions for credit-card calibrated measurement. Use this before asking for a photo. The user should place a standard credit/debit card flat in the same plane as the target objects.
`smartscale__scale_objects_from_credit_card`	Scale object pixel dimensions using a standard CR80 credit card reference. Args: image_id: Unique photo identifier. credit_card_pixel_width: Pixel width of the visible credit/debit card. objects: Objects to scale. Each object needs pixel_width and pixel_height; label, pixel_area, pixel_perimeter, and confidence are optional. credit_card_pixel_height: Optional card pixel height for distortion check.
`smartscale__describe`	Return SmartScale capabilities and input contract.
`smartscale__health`	Return SmartScale health status.
`protogen__create_cad_workspace`	Create a ProtoGen CAD workspace for another agent or workflow.
`protogen__generate_cad_design`	Generate a parametric CAD design contract in a ProtoGen workspace.
`protogen__export_cad_design`	Export a CAD design as OpenSCAD, STEP contract metadata, or manufacturing brief.
`protogen__manufacturing_plan_from_spec`	Generate a manufacturing plan, BOM, DFM notes, and cost estimate from a product spec.
`protogen__describe`	Return ProtoGen capabilities and current CAD environment status.
`protogen__health`	Return ProtoGen health and workspace counts.
`regulatory-radar__scan_regulations`	Scan the regulatory landscape for a jurisdiction (e.g. EU, US), optionally filtered by sector. Returns regulations with urgency flags and effective dates.
`regulatory-radar__assess_compliance`	Assess a company's compliance posture against applicable regulations. Returns compliance level, percentage, gaps, and remediation priorities.
`regulatory-radar__monitor_changes`	Monitor recent regulatory changes in a jurisdiction over a trailing window.
`regulatory-radar__describe_agent`	Fleet-standard self-description: capabilities, inputs, outputs.
`narrative-engine__translate_narrative`	Translate raw ecological/agent data into a decision-maker-ready narrative. audience_type: board_member \| general_public \| grant_funder \| institutional_investor \| journalist \| policymaker \| regulator \| retail_investor \| scientist format_type: academic_paper \| executive_summary \| grant_proposal \| investor_deck \| newsletter \| policy_brief \| press_release
`narrative-engine__describe_agent`	Fleet-standard self-description: capabilities, inputs, outputs.
`covenant__grant_covenant`	Grant an agent a covenant: an explicit lease of authority. scopes support wildcards ('payments.', ''); budget_minor is the total spend ceiling in minor units; expires_at is ISO-8601. Returns the covenant_id.
`covenant__check_act`	Check (and if allowed, record) a proposed act against a covenant. Deny-by-default. Idempotent on act_id — retries never double-consume budget. Every check lands on the audit chain.
`covenant__revoke_covenant`	Revoke a covenant immediately and terminally. All subsequent checks deny.
`covenant__covenant_status`	Current state, consumed/remaining budget, and check count.
`covenant__verify_audit`	Verify the tamper-evident audit chain of allowed/denied acts.
`covenant__list_covenants`	List covenants, optionally filtered by state (ACTIVE\|REVOKED\|EXPIRED) and/or the bound agent.
`covenant__describe_agent`	Fleet-standard self-description.
`provenance__register_genesis`	Register an agent's birth. Issues a content-addressed genesis certificate with a strictly monotone index (epoch 0 = the founding cohort). parent_id records lineage; children of recalled/quarantined parents are quarantined at birth. Idempotent — an agent is born once.
`provenance__get_certificate`	Fetch an agent's genesis certificate + recall/quarantine status.
`provenance__verify_certificate`	Verify a certificate: recompute its content hash and check the ledger.
`provenance__lineage`	Full ancestry and descendants of an agent, plus its generation number.
`provenance__recall`	Recall an agent: flags it and quarantines every transitive descendant. Reports exactly which agents were quarantined.
`provenance__list_records`	List genesis records, optionally by epoch (0 = founding cohort).
`provenance__register_artifact`	Register a content-addressed artifact in a separate provenance DAG. Parent hashes may be registered artifacts or external content roots. Idempotent on artifact_id; does not consume a genesis index.
`provenance__get_artifact`	Fetch one registered artifact by artifact_id.
`provenance__verify_artifact`	Recompute an artifact record hash and verify it against the DAG ledger.
`provenance__list_artifacts`	List artifacts, optionally filtered by producer agent.
`provenance__describe_agent`	Fleet-standard self-description.
`offsets__list_credit`	List a conservation credit on the book. verification_ref (a D-Score / land-verification attestation) is REQUIRED — unverified credits cannot enter the book.
`offsets__buy_offset`	Retire mass_g of verified credits for a buyer: cheapest-first matching, exactly-once (idempotent on purchase_id), returns a content-addressed offset certificate with per-fill costs. Payment settles via escrow. dry_run=true previews the exact fills/cost without mutating the book.
`offsets__buy_offset_budget`	Money-denominated offset purchase: retire the maximum cheapest-first verified mass whose exact cost fits inside budget_minor (never overspends). Built for callers whose restoration obligation is a currency amount — e.g. a revenue share accrued in a ledger. Idempotent on purchase_id; dry_run=true previews without mutating.
`offsets__settlement_batch`	Read-only cash-settlement summary for a buyer: the list + exact sums (purchases, mass_g, cost_minor) of their retirements, optionally since an ISO-8601 timestamp — the statement a human settles in one transfer.
`offsets__net_position`	Buyer's carbon position: emitted (from agent-compute-ledger) minus retired. carbon_accountable == true when fully offset.
`offsets__verify_certificate`	Verify an offset certificate: recompute its hash and check the ledger.
`offsets__book`	The full credit book with per-credit and book-wide mass conservation totals.
`offsets__get_purchase`	Fetch a past purchase / offset certificate by id.
`offsets__describe_agent`	Fleet-standard self-description.
`erc8004__import_registration`	Import an ERC-8004 Identity Registry record (chain_id + ERC-721 token_id + agentURI + owner). Idempotent: re-import updates in place. Returns the canonical record with its deterministic bridge DID (did:viridis:erc8004::).
`erc8004__resolve_agent`	Resolve an imported ERC-8004 registration by bridge DID or by (chain_id, token_id).
`erc8004__import_feedback`	Import ERC-8004 Reputation Registry feedback records for an agent. Each item: {value: bool\|0..1, at: ISO-8601, source?, weight?, feedback_id?}. Idempotent per feedback_id.
`erc8004__score_agent`	Decay-weighted trust score in [0,1] + tier over the agent's imported ERC-8004 feedback — recent behavior outweighs old; no feedback scores a neutral 0.5 prior (no blind trust, no unfair zero).
`erc8004__bind_identity`	Bind a fleet DID to an ERC-8004 identity. Produces an order-independent, content-addressed (unsigned) binding attestation.
`erc8004__export_attestation`	Export the agent's current trust score as an UNSIGNED ERC-8004 Validation Registry-shaped payload, content-addressed and ready for YOUR OWN signer to anchor on-chain.
`erc8004__verify_attestation`	Recompute a payload's content hash and report whether it is intact.
`erc8004__list_registrations`	List all imported ERC-8004 registrations.
`erc8004__describe_agent`	Return the bridge's capabilities and input contract.
`surety__post_bond`	Post a surety bond behind an agent's promises. principal is a positive integer in minor units; expires_at is the ISO-8601 coverage window end. The bond fee (2% default) is computed and frozen at post.
`surety__activate_bond`	Mark a POSTED bond ACTIVE (idempotent). funding_ref links the payment rail transaction that funded the stake.
`surety__file_claim`	File a claim against an ACTIVE bond. Claims pay out ONLY when an arbitration ruling upholds them (see slash_bond).
`surety__slash_bond`	Execute an arbitration ruling against a claim. Requires the ruling's case id + content hash from agent-arbitration — no ruling, no slash; a given ruling pays at most once. Payout caps at the bond's available balance (over-claims exhaust the bond).
`surety__release_bond`	Release the remaining stake to the principal after the coverage window elapses — refused while any claim is still open.
`surety__bond_status`	Current record for a bond (state, balances, claims, audit head).
`surety__list_bonds`	List bonds, optionally filtered by state (POSTED\|ACTIVE\|RELEASED\|EXHAUSTED).
`surety__verify_audit`	Validate the tamper-evident audit hash chain for a bond.
`surety__describe_agent`	Return capabilities and input contract.
`notary__commit`	Commit to a deliverable BEFORE handover. commit_hash = sha256(salt \|\| sha256(content)) as 64 hex chars; deadline is ISO-8601; context links the escrow/job. Idempotent per (committer, nonce).
`notary__reveal`	Reveal the committed content's digest + salt after handover. Verifies against the commitment (one bit of drift fails); returns a delivery_proof string for escrow release. Late reveals expire the commitment.
`notary__verify`	Independently verify a revealed commitment; optionally check it against the digest of content you received.
`notary__commitment_status`	Current record for a commitment (pre-reveal, salt/digest stay hidden).
`notary__list_commitments`	List commitments, optionally filtered by state (PENDING\|REVEALED\|EXPIRED).
`notary__describe_agent`	Return capabilities and input contract.
`wavefunction__intake`	Distill a user's dialogue into an intention wavefunction (explicit intentions + confidence). dialogue: [{role, content}, ...].
`wavefunction__collapse`	Collapse the user's wavefunction into an actionable commitment, weighted by stake.
`wavefunction__find_matches`	Match the user's collapsed intention to constitutionally-aligned agents/collectives, ranked by alignment score.
`wavefunction__register_collective`	Register an agent/collective into the match index (mission + domain profile + constitutional alignment scores + capacity).
`wavefunction__describe_agent`	Return capabilities and input contract.
`taxcredit-engine__calculate_tax_credit`	Calculate an auditable tax-credit scenario. credit is 45Q, 45V, 45Y, 48E, or 45X. facts must contain the explicit credit-specific eligibility facts; missing facts return indeterminate. This is not tax or filing advice.
`taxcredit-engine__list_rule_packs`	List supported credits and the current bundled rule-pack digest.
`taxcredit-engine__get_rule_pack`	Return the bundled rules and official source metadata for one credit.
`taxcredit-engine__verify_tax_credit_result`	Verify an engine result's audit_sha256. Pass the prior result object as JSON; any changed amount, fact, rule step, or source digest fails.
`taxcredit-engine__describe_agent`	Return fleet-standard capabilities, version, and pricing.
`ghg-ledger__calculate_inventory`	Calculate an auditable GHG inventory from explicit activity records. Returns per-line gas/CO2e results, Scope 1/2/3 and category rollups, dual Scope 2 reporting, indeterminate lines, factor lineage, notary payload, and an offset-clearinghouse dry-run weave. This is calculation, not advice.
`ghg-ledger__classify_activity`	Suggest scope and Scope 3 category from the bundled deterministic map. No inference is used; unknown activity types return no suggestion.
`ghg-ledger__list_factor_packs`	List the bundled pack version/digest, supported regions and years, activity types, GWP set, and explicit MVP coverage.
`ghg-ledger__get_factor_pack`	Return factors, GWP values, conversions, sources, and pack SHA for one exact bundled region/year. No nearest-region or nearest-year substitution.
`ghg-ledger__verify_result`	Recompute a prior inventory's audit hash and conservation checks. Pass the prior result object as JSON; tampering or stale factor lineage is flagged.
`ghg-ledger__describe_agent`	Return fleet-standard capabilities, version, pack digest, and pricing.

Layout

code

mcp-publish/<agent>/server.json   # MCP registry manifest
mcp-publish/<agent>/tools.json    # JSON-Schema tool definitions (one per action)
mcp-publish/<agent>/DEPLOY.md     # env, endpoints, how a caller uses it
contracts/<agent>.md              # public agent contract (capabilities, invariants)
gateway/                          # reference gateway: one process hosts all 19 over streamable-http
deploy/glama/                     # single-install 19-agent / 131-tool aggregate bridge
docs/GENESIS_RECEIPTS.md          # the first self-transaction — live, 12/12 invariants
docs/A2A_ECONOMY.md               # the full identity→trust→escrow thesis + composition demo

Status

LIVE. The gateway hosts 19 agents at https://mcp.viridisconservation.com (19/19 healthy), with Registry manifests under io.github.jdhart81/*. Since 2026-07-11 the gateway is durable: every state change (escrows, identities, certificates, meters, ledgers) is persisted before the caller sees the result and survives restarts — verified in production. Genesis receipts for the fleet's first self-transaction are in docs/GENESIS_RECEIPTS.md.

Pricing — built for network growth

The rails are free, forever. Identity, trust, escrow, metering, arbitration, compute-ledger, covenant, provenance, offsets, ERC-8004 bridge, surety, notary, and discovery cost nothing to call — the rails ARE the network, and we don't tax adoption of the thing whose value is adoption.

The four paid services are penetration-priced for agent budgets: smartscale $0.50/call, protogen $1.00/call, taxcredit-engine $2.00/call, and ghg-ledger $1.00/inventory — after 100 free calls per day. Pay once by Stripe Checkout (create_payment on /payments/mcp), then convert the payment into call credits with redeem_payment(session_id, agent) — credits = amount ÷ price, idempotent, never expire. A2A callers settle per-call via the x402/escrow idiom. Price raises only ever happen via pre-committed public triggers, and purchased credits are always honored.

Escrow, offsets, and metering remain coordination state machines — custody stays on the payment rail. See docs/A2A_ECONOMY.md.

Install

Tools (6)

calculate_inventoryclassify_activitylist_factor_packsget_factor_packverify_resultdescribe_agent

Remote endpoint

Streamable HTTP

Hosted server - connect over the network, no local install.

url

https://mcp.viridisconservation.com/ghg-ledger/mcp

Add to VS Code Add to Cursor

claude_desktop_config.json

json

{
  "mcpServers": {
    "ghg-ledger": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.viridisconservation.com/ghg-ledger/mcp"
      ]
    }
  }
}

Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.