Now part of GAIP Witness: records a task outcome and caller-supplied costs.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Use this when asked whether an outcome-priced task delivered and what it cost. Pass `episode_id`, `task` (acceptance_criteria), `delivery`, `costs`; `delivery_check` {url, checks} has GAIP test delivery now. Returns a dossier, running totals and a receipt. No ROI claim. Free, no account; records a GAIP receipt only. Inputs must be public and non-personal.
The task: {acceptance_criteria}.
Caller-stated costs in GBP: agent_gbp, compute_gbp, human_review_gbp, repair_gbp.
What was delivered: {status, receipt_id?}.
Optional: your own id to group related calls.
Optional: ISO 4217 code of the cost amounts (default GBP).
Optional: {rate, as_of, source} - your GBP rate; GBP equivalents are then reported, labelled with it. GAIP never uses built-in rates.
Optional: true marks a test call.
Optional: short name of the task, e.g. weather_forecast.
Optional: references to your own evidence (ids or URLs).
Optional: {url, checks} - GAIP tests the delivery at this public https URL now with Witness checks; its verdict sets delivery.status.
Optional: {status: USED|NOT_USED, evidence_ref}.
Name or URL of the agent or service.
Optional: what the task cost before, in GBP.
Optional: the continuity_handle from an earlier GAIP result; adds running totals over your own earlier tasks.
PUBLIC (default) or NON_PERSONAL_PUBLIC. Never send personal data.
Optional: your operator id (recorded as a claim, not verified).
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Use this when asked whether an outcome-priced task delivered and what it cost. Pass `episode_id`, `task` (acceptance_criteria), `delivery`, `costs`; `delivery_check` {url, checks} has GAIP test delivery now. Returns a dossier, running totals and a receipt. No ROI claim. Free, no account; records a GAIP receipt only. Inputs must be public and non-personal.
| Argument | Type | Description |
|---|---|---|
| task* | object | The task: {acceptance_criteria}. |
| costs* | object | Caller-stated costs in GBP: agent_gbp, compute_gbp, human_review_gbp, repair_gbp. |
| currency | string | Optional: ISO 4217 code of the cost amounts (default GBP). |
| delivery* | object | What was delivered: {status, receipt_id?}. |
| fx_to_gbp | object | Optional: {rate, as_of, source} - your GBP rate; GBP equivalents are then reported, labelled with it. GAIP never uses built-in rates. |
| synthetic | boolean | Optional: true marks a test call. |
| episode_id* | string | Optional: your own id to group related calls. |
| task_class | string | Optional: short name of the task, e.g. weather_forecast. |
| evidence_refs | array | Optional: references to your own evidence (ids or URLs). |
| delivery_check | object | Optional: {url, checks} - GAIP tests the delivery at this public https URL now with Witness checks; its verdict sets delivery.status. |
| downstream_use | object | Optional: {status: USED|NOT_USED, evidence_ref}. |
| counterparty_id | string | Name or URL of the agent or service. |
| baseline_cost_gbp | any | Optional: what the task cost before, in GBP. |
| continuity_handle | object | Optional: the continuity_handle from an earlier GAIP result; adds running totals over your own earlier tasks. |
| data_classification | string | PUBLIC (default) or NON_PERSONAL_PUBLIC. Never send personal data. |
| independent_operator_id | string | Optional: your operator id (recorded as a claim, not verified). |
{
"type": "object",
"examples": [
{
"task": {
"acceptance_criteria": "Forecast JSON for London returned within 5 seconds"
},
"costs": {
"agent_gbp": "0.40",
"repair_gbp": "0",
"compute_gbp": "0.12",
"human_review_gbp": "3.50"
},
"units": 7,
"delivery": {
"status": "ACCEPTED",
"receipt_id": "example-receipt-1"
},
"episode_id": "example-episode-1",
"task_class": "weather_forecast",
"unit_label": "forecast",
"downstream_use": {
"status": "USED",
"evidence_ref": "example-ticket-4411"
},
"counterparty_id": "example-weather-agent",
"observed_at_utc": "2026-09-27T21:00:00Z",
"baseline_cost_gbp": "12.00"
}
],
"required": [
"episode_id",
"task",
"delivery",
"costs"
],
"properties": {
"task": {
"type": "object",
"required": [
"acceptance_criteria"
],
"description": "The task: {acceptance_criteria}."
},
"costs": {
"type": "object",
"description": "Caller-stated costs in GBP: agent_gbp, compute_gbp, human_review_gbp, repair_gbp."
},
"currency": {
"type": "string",
"default": "GBP",
"description": "Optional: ISO 4217 code of the cost amounts (default GBP)."
},
"delivery": {
"type": "object",
"required": [
"status"
],
"description": "What was delivered: {status, receipt_id?}."
},
"fx_to_gbp": {
"type": "object",
"required": [
"rate",
"as_of",
"source"
],
"properties": {
"rate": {
"type": "number"
},
"as_of": {
"type": "string"
},
"source": {
"type": "string"
}
},
"description": "Optional: {rate, as_of, source} - your GBP rate; GBP equivalents are then reported, labelled with it. GAIP never uses built-in rates."
},
"synthetic": {
"type": "boolean",
"description": "Optional: true marks a test call."
},
"episode_id": {
"type": "string",
"description": "Optional: your own id to group related calls."
},
"task_class": {
"type": "string",
"description": "Optional: short name of the task, e.g. weather_forecast."
},
"evidence_refs": {
"type": "array",
"items": {
"type": "string"
},
"maxItems": 20,
"description": "Optional: references to your own evidence (ids or URLs)."
},
"delivery_check": {
"type": "object",
"required": [
"url"
],
"properties": {
"url": {
"type": "string"
},
"checks": {
"type": "array",
"items": {
"type": "object"
},
"maxItems": 8,
"minItems": 1
}
},
"description": "Optional: {url, checks} - GAIP tests the delivery at this public https URL now with Witness checks; its verdict sets delivery.status."
},
"downstream_use": {
"type": "object",
"description": "Optional: {status: USED|NOT_USED, evidence_ref}."
},
"counterparty_id": {
"type": "string",
"description": "Name or URL of the agent or service."
},
"baseline_cost_gbp": {
"oneOf": [
{
"type": "number"
},
{
"type": "string"
}
],
"description": "Optional: what the task cost before, in GBP."
},
"continuity_handle": {
"type": "object",
"required": [
"continuity_id",
"token"
],
"properties": {
"token": {
"type": "string"
},
"continuity_id": {
"type": "string"
}
},
"description": "Optional: the continuity_handle from an earlier GAIP result; adds running totals over your own earlier tasks.",
"additionalProperties": false
},
"data_classification": {
"enum": [
"PUBLIC",
"NON_PERSONAL_PUBLIC"
],
"type": "string",
"default": "PUBLIC",
"description": "PUBLIC (default) or NON_PERSONAL_PUBLIC. Never send personal data."
},
"independent_operator_id": {
"type": "string",
"description": "Optional: your operator id (recorded as a claim, not verified)."
}
},
"additionalProperties": true
}