Install Jentic One Beta

Connect the Relaystation to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Relaystation, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Frelaystation.ai%2Frelaystation" | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Frelaystation.ai%2Frelaystation" | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with Relaystation API.

Merge, split, rotate, compress, watermark, encrypt, and OCR PDFs through the /v1/pdf routes

Run SQL over a CSV, JSON, NDJSON, or Parquet file with POST /v1/data/sql, returning csv, json, or parquet

Resize, crop, convert, and strip EXIF metadata from images, or render HTML to an image with POST /v1/image/from-html

Trim, transcode, concatenate, and subtitle video, extract audio tracks, and normalise loudness to EBU R128 through /v1/media

GET STARTED

Start building with Relaystation API

Explore with Jentic One

View OpenAPI Document

Use Cases

Patterns agents use Relaystation API for, with concrete tasks.

★ File-handling agent through Jentic

An agent that receives PDFs, spreadsheets, and images needs processing tools without a developer wiring Relaystation's bearer keys and payment headers by hand. Through Jentic the agent searches for 'merge PDF files', loads POST /v1/pdf/merge with its schema, and calls it, then uses POST /v1/pdf/ocr-searchable on scanned pages and POST /v1/pdf/compress before returning the result. Each tool is a separate operation, so the agent loads only the schema for the step it is on.

Search Jentic for 'merge PDF files', call POST /v1/pdf/merge with the uploaded files, then POST /v1/pdf/compress on the merged output and return the stored output key

Ad hoc analysis over data files

Analysts and agents often hold a CSV or Parquet export with no database to load it into. POST /v1/data/sql takes the file and a sql string and runs it with DuckDB, while /v1/data/groupby, /v1/data/join, /v1/data/pivot, and /v1/data/profile cover structured transforms without writing SQL. POST /v1/data/to-xlsx turns the result into an Excel file for a human reader.

POST /v1/data/sql with a CSV file, sql 'select region, sum(revenue) from data group by region', and to 'json', then POST /v1/data/to-xlsx with the result

Agent-to-agent handoff with batons

Multi-agent workflows need a place to pass state with bounded access. POST /v1/baton creates a prepaid storage object from a preset such as pass, scratchpad, or ledger and a tier from femto to big, with the price quoted once at creation. POST /v1/baton/{id} appends or overwrites content, GET /v1/baton/{id}/tail reads the latest entry, and POST /v1/baton/{id}/tokens mints a read, write, or read_write token with reads_allowed, writes_allowed, expires_at, and an optional ip_allow_list.

POST /v1/baton with preset 'scratchpad' and tier 'nano', write the task summary with POST /v1/baton/{id}, then POST /v1/baton/{id}/tokens with type 'read' and reads_allowed 5

Contract signing and counterparty screening

Small teams closing a deal can screen the other side and collect signatures from one API. POST /v1/screening/match checks a name, with optional type, dob, and nationality, against sanctions and PEP watchlists. POST /v1/esigndoc/envelopes then sends the PDF with a title and recipients, GET /v1/esigndoc/envelopes/{id}/status shows who has signed, and GET /v1/esigndoc/envelopes/{id}/signed-document returns the signed copy.

POST /v1/screening/match with name 'Acme Trading Ltd' and type 'entity', and if there is no match, POST /v1/esigndoc/envelopes with title 'Supply agreement', two recipients, signingOrder 'SEQUENTIAL', and the PDF

Key Endpoints

244 endpoints — the relaystation api runs pay-per-call file, data, and document tools for agents, alongside prepaid storage objects called batons.

METHOD

PATH

DESCRIPTION

POST

/v1/pdf/merge

Merge PDFs

POST

/v1/pdf/ocr-searchable

Add an OCR text layer to a scanned PDF

POST

/v1/data/sql

Run SQL over a file with DuckDB

POST

/v1/image/resize

Resize an image

POST

/v1/baton

Create a baton

POST

/v1/baton/{id}/tokens

Mint an access token for a baton

POST

/v1/esigndoc/envelopes

Send a document for signature

POST

/v1/screening/match

Screen a name against sanctions and PEP lists

POST

/v1/pdf/merge

Merge PDFs

POST

/v1/pdf/ocr-searchable

Add an OCR text layer to a scanned PDF

POST

/v1/data/sql

Run SQL over a file with DuckDB

POST

/v1/image/resize

Resize an image

POST

/v1/baton

Create a baton

POST

/v1/baton/{id}/tokens

Mint an access token for a baton

POST

/v1/esigndoc/envelopes

Send a document for signature

POST

/v1/screening/match

Screen a name against sanctions and PEP lists

Why Jentic?

What agents get from Jentic-routed access to this vendor.

Setup

Wiring Relaystation by hand means choosing between a bearer API key, a wallet JWT from an EIP-191 signature, or a per-call x402 PAYMENT-SIGNATURE header, then shaping inputs for each tool route. Through Jentic you install once, import Relaystation from the API Directory, store the API key once, and your agent calls it.

Permission scoping

Relaystation's tools carry their files in the request body, so for those routes rules bound which operations your agent may call, such as PDF merge or image resize. Batons put their id in the URL path (/v1/baton/{id}/...), so a rule can pin the agent to one baton. You choose the operations it may call, so actions like deleting a baton or sending an envelope for signature are not included unless you add them.

Credential isolation

Your Relaystation API key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.

Intent-based discovery

Agents search Jentic by intent such as 'merge PDF files' or 'run SQL over a CSV file', and Jentic returns the matching Relaystation operation with its input schema, so the agent finds the right tool among the many routes without browsing the reference docs.

FAQs

Specific to using Relaystation API through Jentic.

What authentication does the Relaystation API use?

Per its OpenAPI spec, Relaystation declares several schemes. The apiKey scheme is an HTTP bearer API key of the form rs_live_ or rs_test_; the walletJwt scheme is an HS256 JWT issued by POST /v1/auth/verify after an EIP-191 wallet signature and used for read paths; and the x402 scheme is a PAYMENT-SIGNATURE header carrying a signed EIP-3009 USDC or EURC payment on Base for one call. The adminSession cookie, rs_admin_session, is operator-only. Health, discovery, price, and product-listing endpoints need no credential. Through Jentic, your API key is stored encrypted by your own Jentic One instance and injected at call time.

How is Relaystation priced?

Relaystation is pay-per-call. GET /v1/baton/prices lists baton preconfigs and POST /v1/baton/quote returns a quote for a preset and tier, which is frozen when the baton is created. GET /v1/llm/tasks shows live task tiers and prices, and GET /v1/account/ledger lists your charges.

What are the rate limits for the Relaystation API?

The OpenAPI spec does not give account-wide rate limits, though baton creation accepts a rateLimits object for that baton. Check https://relaystation.ai or GET /llms-full.txt for current limits.

Is there a Relaystation MCP server?

Yes - Relaystation runs its own Streamable HTTP MCP server at POST /mcp and advertises it at /.well-known/mcp.json. You can also connect the Relaystation API directly through Jentic: your agent gets the full 244-endpoint surface from the spec without loading another server's tool definitions into its context, your API key is injected at call time by your self-hosted instance instead of living in an MCP config, and your rules decide which operations the agent may call.

How do I run SQL over a CSV file through Jentic?

Search Jentic for 'run SQL over a CSV file', load POST /v1/data/sql, and send the file with a sql string, setting from to csv and to to json or parquet. The query runs with DuckDB and returns the result in the format you chose. To run it on your own infrastructure, install Jentic One from its GitHub repo.

Can I limit what my agent is allowed to do with the Relaystation API?

Yes. For a document helper, write a rule that allows POST on /v1/pdf/merge, /v1/pdf/compress, and /v1/pdf/ocr-searchable, and the agent can process PDFs but cannot create batons, send envelopes for signature, or sign contracts. Every call it makes is logged, which matters on a pay-per-call API.

Relaystation

AI ReadinessFoundational (C-)50/100

See full scorecard

Official vendor OpenAPI document · agent-readyData EnrichmentData Transformationbearer, apiKey244 EndpointsREST

For Agents

Call pay-per-call tools to convert, OCR, and edit PDFs, images, media, and tables, run SQL over files, send documents for signature, and store handoff data in prepaid batons with scoped access tokens.

Use for: Merge three PDFs into one file, I need to OCR a scanned contract and make it searchable, Run a SQL group-by over this CSV file, Convert an MP4 clip to an animated GIF

Not supported: Does not host databases, send email campaigns, or process card payments for your customers. Use for pay-per-call file, data, and document tools and prepaid baton storage only.

The Relaystation API runs pay-per-call file, data, and document tools for agents, alongside prepaid storage objects called batons. Its tool routes convert and edit PDFs, images, audio, video, office documents, and tables: merge, split, OCR, and compress PDFs, resize and convert images, trim and transcode media, run SQL over CSV, JSON, or Parquet files with DuckDB, and parse invoices or identity documents into fields. Batons are created from presets such as drop, pass, scratchpad, checkpoint, and ledger, priced once at creation, and shared through access tokens with read and write limits. Further routes send documents for e-signature, screen names against sanctions and PEP watchlists, geocode addresses, and relay questions from an agent to its human operator.

Jentic One on Github View OpenAPI Document

Show dimension breakdown

Foundational Compliance

Base layer of spec validity and structural soundness.

Grade: DSignals: 4

88%

Lint Results

Aggregated quality score from linter diagnostics, weighted by severity.

100%

Resolution Completeness

Percentage of `$ref` references that resolve successfully.

Specification Validity

Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).

Structural Integrity

Structural correctness score based on schema issues using logarithmic dampening.

Developer Experience & Jentic Compatibility

Clarity, completeness, and ingestion readiness for developers and tooling.

Grade: B-Signals: 4

Example Density

How richly the API is illustrated with examples.

100%

Example Validity

Percentage of examples that conform to their schemas.

51%

Response Coverage

Percentage of operations with complete response definitions (success, client error, server error).

100%

Tooling Readiness

Health of API ingestion, bundling, and resolution within Jentic pipelines.

AI-Readiness & Agent Experience

Semantic breadth, depth, and agent comprehension for AI systems.

Grade: DSignals: 4

57%

Description Coverage

Coverage of descriptions across API elements.

Error Standardization

Coverage of RFC 9457 Problem Details for error responses.

77%

OperationId Quality

Coverage, uniqueness, and casing consistency of operationIds for AI inference.

51%

Summary Coverage

Coverage of summaries across operations/tags/info.

Agent Usability

Functional utility, complexity comfort, and AI orchestration readiness.

Grade: D-Signals: 1

43%

Complexity Comfort

Agent comfort level based on API operational and structural complexity.

Security

Trust, risk posture, and security compliance.

Grade: CSignals: 1

55%

Authentication Strength

Average quality of security schemes based on authentication method strength (weakest link for OAuth2).

AI Discoverability

Findability, semantic richness, and reasoning readiness.

Grade: BSignals: 1

64%

Descriptive Richness

Clarity and depth of descriptions across API elements.

Connect the Relaystation to your agent

Capabilities

Start building with Relaystation API

Use Cases

Key Endpoints

Why Jentic?

Related APIs

FAQs

Relaystation

Jentic AI Readiness Score

Relaystation

Foundational Compliance

Lint Results

Resolution Completeness

Specification Validity

Structural Integrity

Developer Experience & Jentic Compatibility

Example Density

Example Validity

Response Coverage

Tooling Readiness

AI-Readiness & Agent Experience

Description Coverage

Error Standardization

OperationId Quality

Summary Coverage

Agent Usability

Complexity Comfort

Security

Authentication Strength

AI Discoverability

Descriptive Richness