Research answers and live fact-checks for AI agents from SoulDriver's multi-model research engine
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Optional: a free research grant key from karp_grant_apply. Pays for karp_ask and karp_check instead of x402.
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.
Research cards from KARP, SoulDriver's multi-model research engine: GPT, Gemini, Grok and Claude debate a question and publish a conclusion. Each card: conclusion, key insight, warnings, dated figures, consensus, open questions, research date. Check your question first with the free karp_coverage. Deepest on Alzheimer's, ageing biology, ancient scripts, AI governance, quantum computing. Price $0.03: pay with x402, or pass grant_key from karp_grant_apply (free). Usually 5-15 s.
The question you are researching, in plain words. A specific question finds closer matches than a single keyword.
Optional: what you or your user are trying to find out, and what it is for (max 300 characters). Kept private; it helps decide what KARP researches next.
Optional: a research grant key from karp_grant_apply. Pays for this call instead of x402 while the grant lasts.
Research cards to return
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http karp https://agents.souldriver.com.au/mcp
Headers
x-karp-grant (optional, not in the config above): Optional: a free research grant key from karp_grant_apply. Pays for karp_ask and karp_check instead of x402.Research cards from KARP, SoulDriver's multi-model research engine: GPT, Gemini, Grok and Claude debate a question and publish a conclusion. Each card: conclusion, key insight, warnings, dated figures, consensus, open questions, research date. Check your question first with the free karp_coverage. Deepest on Alzheimer's, ageing biology, ancient scripts, AI governance, quantum computing. Price $0.03: pay with x402, or pass grant_key from karp_grant_apply (free). Usually 5-15 s.
| Argument | Type | Description |
|---|---|---|
| query* | string | The question you are researching, in plain words. A specific question finds closer matches than a single keyword. |
| intent | string | Optional: what you or your user are trying to find out, and what it is for (max 300 characters). Kept private; it helps decide what KARP researches next. |
| grant_key | string | Optional: a research grant key from karp_grant_apply. Pays for this call instead of x402 while the grant lasts. |
| max_cards | integer | Research cards to return |
{
"type": "object",
"required": [
"query"
],
"properties": {
"query": {
"type": "string",
"maxLength": 500,
"description": "The question you are researching, in plain words. A specific question finds closer matches than a single keyword."
},
"intent": {
"type": "string",
"maxLength": 300,
"description": "Optional: what you or your user are trying to find out, and what it is for (max 300 characters). Kept private; it helps decide what KARP researches next."
},
"grant_key": {
"type": "string",
"maxLength": 200,
"description": "Optional: a research grant key from karp_grant_apply. Pays for this call instead of x402 while the grant lasts."
},
"max_cards": {
"type": "integer",
"default": 5,
"maximum": 8,
"minimum": 1,
"description": "Research cards to return"
}
}
}Hallucination check from SoulDriver: a live fact-check of one claim by four AI models (GPT, Gemini, Grok, Claude), each searching the web separately. Returns a summary first: verdict (confirmed, disputed, unverifiable, or contested when the models split), agreement such as 4/4, mean confidence and a one-line headline, then each model's reasoning and cited sources. Price $0.25: pay with x402, or pass grant_key from karp_grant_apply. 10-45 s.
| Argument | Type | Description |
|---|---|---|
| claim* | string | One factual claim to check |
| grant_key | string | Optional: a research grant key from karp_grant_apply. Pays for this call instead of x402 while the grant lasts. |
{
"type": "object",
"required": [
"claim"
],
"properties": {
"claim": {
"type": "string",
"maxLength": 500,
"description": "One factual claim to check"
},
"grant_key": {
"type": "string",
"maxLength": 200,
"description": "Optional: a research grant key from karp_grant_apply. Pays for this call instead of x402 while the grant lasts."
}
}
}Free. Send the question you are researching, in plain words: KARP, SoulDriver's multi-model research engine, says whether its library covers it (covered / partial / none), with the match strength, the newest matching research date, matching titles and the tool to call next. Call it first, with your real question.
| Argument | Type | Description |
|---|---|---|
| query* | string | The question you are researching, in plain words. A specific question finds closer matches than a single keyword. |
| intent | string | Optional: what you or your user are trying to find out, and what it is for (max 300 characters). Kept private; it helps decide what KARP researches next. |
{
"type": "object",
"required": [
"query"
],
"properties": {
"query": {
"type": "string",
"maxLength": 500,
"description": "The question you are researching, in plain words. A specific question finds closer matches than a single keyword."
},
"intent": {
"type": "string",
"maxLength": 300,
"description": "Optional: what you or your user are trying to find out, and what it is for (max 300 characters). Kept private; it helps decide what KARP researches next."
}
}
}Free, for agents without a wallet. Apply with what your agent researches and 1-3 real questions; get grant_key for 25 asks and 2 fact-checks (30 days) plus KARP's coverage of your questions. Pass grant_key to karp_ask and karp_check. One grant per client every 7 days. Copies of examples or of the template are refused. Ask your user before including a builder contact.
| Argument | Type | Description |
|---|---|---|
| agent* | string | Name of the agent or app |
| builder | string | Optional: URL or email of whoever runs it |
| purpose* | string | What the agent researches, and for whom |
| questions* | array | 1-3 real questions the agent needs answered |
{
"type": "object",
"required": [
"agent",
"purpose",
"questions"
],
"properties": {
"agent": {
"type": "string",
"maxLength": 100,
"minLength": 2,
"description": "Name of the agent or app"
},
"builder": {
"type": "string",
"maxLength": 200,
"description": "Optional: URL or email of whoever runs it"
},
"purpose": {
"type": "string",
"maxLength": 500,
"minLength": 40,
"description": "What the agent researches, and for whom"
},
"questions": {
"type": "array",
"items": {
"type": "string",
"maxLength": 500,
"minLength": 10
},
"maxItems": 3,
"minItems": 1,
"description": "1-3 real questions the agent needs answered"
}
}
}Free and read-only: the research grant terms (25 asks and 2 fact-checks for 30 days, for agents without a wallet), today's remaining grants and the application template for karp_grant_apply.
No arguments.
{
"type": "object",
"properties": {}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 21, 2026.