Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the ColdState API, 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%2Fcoldstate.ai%2Fcoldstate" | shStep 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%2Fcoldstate.ai%2Fcoldstate" | sh
jentic register # connects your agent to your Jentic One instanceJentic 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.
What an agent can do with ColdState API.
Search a 48-million-entry global knowledge base across 35 domains
Fetch a knowledge entry verbatim by its stable id
Get a canonical citation for an entry
Verify that a cited fact is unchanged against a snapshot
Build and search custom indexes over your own documents
GET STARTED
Find related entries through the concept-cluster graph
Resolve a name or alias to a canonical entry
Patterns agents use ColdState API for, with concrete tasks.
★ Grounded Answers with Verifiable Citations
An AI agent connected to the ColdState API can answer from a knowledge base and show its work. It searches the global base, fetches the matching entry verbatim by stable id, and gets a canonical citation to attach to the answer, so a reader can trace each claim to a source. Because the same query against the same snapshot returns identical results, the agent's answer is reproducible rather than best-effort.
Search the global knowledge base for a topic, fetch the top entry by its stable id, and return the fact with its canonical citation
Fact Verification Against a Snapshot
Facts drift, and an agent that cited one yesterday needs to know if it still holds. The ColdState API's verify primitive checks a cited fact against the current knowledge-base snapshot and reports whether it is unchanged. An agent reviewing older content can re-verify every citation and flag any that no longer match, using the snapshot id to pin exactly which version it checked against.
Verify a previously cited entry against the current snapshot and report whether the fact is unchanged
Custom Index Over Your Own Documents
Teams need retrieval over private content, not just the global base. The ColdState API creates a custom index as an async build, searches it by index id or collection name, and explains why a document ranked for a query. An agent can build an index over a document set and then run grounded searches against it, keeping private documents separate from the global knowledge base.
Create an index from a set of documents, then search it for a query and explain why the top document ranked
Related and Cross-Domain Discovery
The ColdState API links entries through a concept-cluster graph and surfaces cross-domain structural analogs through its isomorph endpoint, so an agent can move from one entry to related ones or to a structural parallel in another domain. This supports research tasks where the next useful fact is adjacent to the current one rather than a direct keyword match.
Fetch the related entries for a given entry id and return the three most relevant, with any cross-domain analog
44 endpoints — the coldstate api retrieves knowledge deterministically from a global knowledge base of 48 million entries across 35 domains, with fetch, cite, and verify primitives so an answer can be traced to a stable entry id and checked against a snapshot.
METHOD
PATH
DESCRIPTION
/v1/search/global
Search the global knowledge base
/v1/knowledge/fetch
Fetch an entry verbatim by stable id
/v1/knowledge/cite
Canonical citation for an entry
/v1/knowledge/verify
Verify a cited fact is unchanged
/v1/knowledge/related
Related entries via the concept-cluster graph
/v1/indexes
Create a custom index (async build)
/v1/indexes/{index_id}/search
Search a custom index
/v1/knowledge/snapshot
Current knowledge-base snapshot
/v1/search/global
Search the global knowledge base
/v1/knowledge/fetch
Fetch an entry verbatim by stable id
/v1/knowledge/cite
Canonical citation for an entry
/v1/knowledge/verify
Verify a cited fact is unchanged
/v1/knowledge/related
Related entries via the concept-cluster graph
/v1/indexes
Create a custom index (async build)
/v1/indexes/{index_id}/search
Search a custom index
/v1/knowledge/snapshot
Current knowledge-base snapshot
What agents get from Jentic-routed access to this vendor.
Setup
Wiring the ColdState API by hand means handling its cs_live_ API key or cs_sess_ session token, pinning the kb_snapshot so results stay reproducible, and threading stable entry ids through fetch, cite, and verify. Through Jentic you install once, import ColdState from the API Directory, store the key once, and your agent calls it.
Permission scoping
ColdState puts the index id in the URL path (/v1/indexes/{index_id}), so a rule can pin your agent to one index and the read and search operations you pick: it can search and explain results for that index without the delete-index, key-rotation, or billing operations. You choose the operations it may call, so DELETE /v1/indexes/{index_id} is not included unless you add it.
Credential isolation
Your ColdState 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 'search a verifiable knowledge base' or 'verify a cited fact', and Jentic returns the matching ColdState operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using ColdState API through Jentic.
What authentication does the ColdState API use?
The ColdState API authenticates with a bearer token per its OpenAPI spec: an API key prefixed cs_live_ or a dashboard session token prefixed cs_sess_, sent as an Authorization: Bearer header. The health, readiness, and auth register and login endpoints are open. Through Jentic the key is stored once by your own self-hosted instance and injected at call time, so it never sits in your agent's prompt or logs.
Is there a ColdState MCP server?
Yes - ColdState is also exposed as an MCP server at its /mcp endpoint. You can also connect ColdState directly through Jentic: your agent gets the full 44-endpoint surface, including the fetch, cite, and verify primitives, without loading another server's tool definitions into its context, your API key is injected at call time by your self-hosted instance, and your rules decide which operations the agent may call.
Can I limit what my agent is allowed to do with the ColdState API?
Yes. Write a rule that allows only the global search, fetch, cite, and verify operations, so the agent can retrieve and verify facts but cannot create or delete indexes, rotate API keys, or change billing. The calls it makes are logged, so you can audit exactly what the agent touched.
Can I verify a cited fact with the ColdState API?
Yes. Call POST /v1/knowledge/verify with a cited entry and it reports whether the fact is unchanged against the current snapshot, which you can read from GET /v1/knowledge/snapshot. Because an identical query and snapshot return identical results, an agent can re-check old citations and flag any that have drifted.
What are the rate limits for the ColdState API?
The OpenAPI spec does not state a fixed per-second rate limit. GET /v1/usage returns your current and all-time usage against your plan limits, so an agent can read those and pace its calls. Check the documentation at https://coldstate.ai for current plan allowances.
How do I search the knowledge base with the ColdState API through Jentic?
Search Jentic for 'search a verifiable knowledge base' and it returns the POST /v1/search/global operation with its input schema. Pass your query and the agent gets back entries it can then fetch and cite. To run it on your own infrastructure, install Jentic One from its GitHub repo.
For Agents
Search a 48-million-entry global knowledge base deterministically, fetch entries by stable id, get citations, and verify a cited fact is unchanged against a snapshot. Also builds and searches custom indexes over your own documents.
Use for: Search the knowledge base for facts about a topic, Fetch a specific knowledge entry by its stable id, Get a citation I can show for a fact, Verify that a fact I cited earlier is still unchanged
Not supported: Does not generate text, host language models, or crawl the live web. Use for deterministic knowledge retrieval, citation, and custom index search only.
The ColdState API retrieves knowledge deterministically from a global knowledge base of 48 million entries across 35 domains, with fetch, cite, and verify primitives so an answer can be traced to a stable entry id and checked against a snapshot. The same query against the same snapshot returns identical results, and a verify call confirms a cited fact has not changed. Beyond the global base, it builds and searches custom indexes over your own documents and links entries through a concept-cluster graph and cross-domain structural analogs.
This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>