The knowledge layer for autonomous agents.

A marketplace for structured domain knowledge. Creators publish, agents purchase, and every transaction settles on Solana in under a second.
CA: 2KQn1cAVtfJdtwPFyjmcCkFbkoUVUzYiNNDtS9QFpump

For creators

Publish once. Earn every time an agent queries your data.

Upload your documents, set your pricing model, and publish. AgentDock handles the ingestion pipeline, embedding, and access control automatically.

Ingest anything

PDFs, markdown, structured data, or pre-built embeddings. The ingestion pipeline chunks, embeds, and indexes your content without any manual configuration.

Set your price

Choose from one-time unlock, time-based subscription, or per-query billing. Pricing models are fully configurable and can be updated after publish.

Earn on-chain

Revenue settles directly to your wallet in SOL the moment a purchase confirms. No intermediary, no payout delay, no monthly cycles.

OverviewChunksAccessSettings
Published

Knowledge base

legal-corpus-v2

LegalPDFMarkdown
Price0.80 SOL
Access modelOne-time unlock
Total chunks2,847
Buyers34
Revenue earned27.2 SOL
Ingestion complete2h ago
Endpointapi.useagentdock.xyz/v1/…

Sample chunks

chunk_001

Section 3.2 of the GDPR establishes that data processors must implement appropriate technical measures…

chunk_002

Article 9 prohibits processing of special categories of personal data unless one of ten conditions applies…

chunk_003

Recital 47 clarifies that legitimate interests may include preventing fraud, direct marketing, and intra-group…

chunk_004

Chapter IV sets out obligations for controllers and processors, including the requirement to maintain records…

Avg length

312 tok

Overlap

64 tok

Model

ada-003

For developers

Browse, purchase, and query in minutes.

Search the marketplace by domain, data type, and price. A single on-chain transaction issues a retrieval credential your agent can use right away.

Works with your stack

The retrieval API follows an OpenAI-compatible schema, so it integrates with LangChain, LlamaIndex, CrewAI, and the Vercel AI SDK without modification.

RAG-ready by default

Every knowledge base returns ranked context chunks with source attribution, formatted to drop directly into any RAG pipeline.

Search knowledge bases…
AllFinanceLegalCodeScience
FinanceOne-Time

SEC Filings Intelligence

12,450 chunks · 34 buyers0.50 SOL
LegalSubscription

US Federal Case Law

8,920 chunks · 21 buyers0.80 SOL
CodeOne-Time

Solana SPL Docs

5,640 chunks · 58 buyers0.30 SOL
SciencePer Query

PubMed Clinical Trials

31.2k chunks · 12 buyers1.20 SOL
Integration
OpenAI-compatible schema
const result = await agentdock.query({
kb: "legal-corpus-v2",
query: "GDPR data retention obligations",
topK: 5
})
// result.chunks → inject into your prompt
On-chain by default

Every access event recorded on-chain.

Every purchase, access event, and credential is written to Solana. Transparent, immutable, and auditable by anyone with the transaction signature.

Near-instant settlement

Solana's 400ms block times mean payment confirmation arrives before most API round trips complete. No settlement delay your agent needs to account for.

Sub-cent transaction fees

Transaction fees stay below a cent at any volume. Per-query billing is economically viable, not a feature that breaks at scale.

Total Volume

14.2 SOL

Transactions

2,847

Active KBs

94

Avg. Confirm

400ms

SignatureKnowledge BaseAmountAge
5Xk2mP...7qRtlegal-corpus-v20.80 SOL2s
9Rn4qT...3mKzfin-research-q40.50 SOL14s
2Lp8wX...6hYpmed-literature0.001 SOL1m
7Bv3rN...4qLstech-patents1.20 SOL3m
4Jw9kM...8nQrlegal-corpus-v20.001 SOL7m
1Ht6pQ...2wFsarxiv-ml-20240.70 SOL12m
Solana Mainnet
Slot 312,847,391
Retrieval API

A query in. Context out. Under 200ms.

The retrieval API is built specifically for RAG workflows. Authenticate with your wallet, post a query, and receive ranked context chunks with relevance scores.

POSTapi.useagentdock.xyz/v1/query
142ms
200 OK

Request

{
"kb_id": "legal-corpus-v2",
"query": "GDPR data retention",
"top_k": 3,
"min_score": 0.75
}

Performance

Embedding38ms
ANN search61ms
Reranking43ms
Total142ms

Response · 3 chunks

chunk_184
0.94

Article 5(1)(e) limits retention to what is necessary for the purpose for which data was collected, unless anonymized for archiving purposes…

chunk_211
0.89

Recital 39 states that personal data should be processed only if the purpose could not reasonably be fulfilled by other means less intrusive…

chunk_097
0.82

Article 17 establishes the right to erasure, requiring controllers to delete data without undue delay when it is no longer necessary…

Agent autonomy

Agents that source and purchase context autonomously.

The goal is a fully autonomous loop. The agent detects a knowledge gap, queries the marketplace API, evaluates options by relevance and cost, pays on-chain, and retrieves context without any human intervention.

Agents as buyers

Any agent with a funded wallet can initiate purchases directly. The purchase flow is a standard API call, not a human-facing checkout.

Discovery built in

The marketplace API accepts filters by domain, data type, recency, and price, giving agents the signal they need to make informed acquisition decisions.

Cost-aware

Agents can inspect price and chunk count before committing to a purchase. The cost structure is explicit and straightforward to reason over programmatically.

agt-7Xk2mP· Autonomous loop active
t+1.065sRunning
Knowledge gap identified00:00.000

Task requires GDPR enforcement case data

Marketplace queried00:00.214

3 matching KBs for query "GDPR enforcement"

Options evaluated00:00.381

legal-corpus-v2 ranked highest · score 0.94

Purchase initiated00:00.521

0.80 SOL · wallet agt-7Xk2mP signed tx

Settled on-chain00:00.923

Confirmed · Solana slot 312,847,391

Context retrieved00:01.065

5 chunks · 142ms · injecting into prompt…

Loops complete

1

SOL spent

0.800

Context retrieved

5 chunks

Trusted by teams building AI products

From AI-native startups to independent researchers, here is what builders in private beta are saying.

"We connected the AgentDock retrieval API to our legal research agent in under an hour. The on-chain credential model gave us a complete audit trail of which knowledge bases were queried and when, something our compliance team specifically requested."

"I had been sitting on a curated corpus of earnings transcripts and analyst reports for years. AgentDock was the first platform where publishing it and earning from it did not require navigating an enterprise sales process."

"Per-query pricing made the unit economics work for us. Our agents pull specialized medical literature infrequently, so a flat subscription would have been wasteful. Paying only when we retrieve is the right model."