canonical: https://jentic.com/apis/bindstack.ai/bindstack

# Bindstack RAG Query API

The Bindstack RAG Query API runs retrieval-augmented queries against a meeting's transcript. Given a meeting_code and a natural-language query, it returns the most relevant transcript passages ranked by similarity, along with the matched meeting title and a total result count. It is a read-only retrieval surface for pulling grounded context out of recorded meetings.

## For AI agents

Run a retrieval-augmented query against a Bindstack meeting transcript and get back the most relevant passages ranked by similarity with the meeting title and result count.

## Scope

Does not create meetings, record audio, or transcribe calls. Use for querying existing Bindstack meeting transcripts only.

## Capabilities

- Run a retrieval-augmented query against a meeting's transcript
- Return the most relevant transcript passages ranked by similarity score
- Read the matched meeting title and total result count for a query

## Use cases

### AI agent transcript retrieval

An AI agent answering questions about past meetings can pull grounded context on demand. Through Jentic the agent searches by intent and calls the query endpoint, receiving ranked transcript passages with similarity scores it can cite rather than guessing from memory.

Example prompt: Use Jentic to search 'query a meeting transcript', call POST `/v1/external/rag/query` with meeting_code 'MTG-TUR7-WGK4' and query 'what did they decide about VS Code', and return the top results.

### Meeting knowledge retrieval

A knowledge assistant surfaces what was said in a specific meeting without replaying the recording. The agent sends a query scoped to one meeting_code and reads the total_results count and the ranked results array to build a summary.

Example prompt: Call POST `/v1/external/rag/query` with a meeting_code and a natural-language query, then read total_results and the ranked results array.

### Grounded answer assembly

When an agent must cite sources, it retrieves passages with their similarity scores and composes an answer anchored to the transcript. This keeps generated answers traceable to the meeting content that supports them.

Example prompt: Send a query for a meeting_code via POST `/v1/external/rag/query` and use the returned passages and similarity scores to compose a cited answer.

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | `/v1/external/rag/query` | Run a retrieval-augmented query against a meeting transcript |

## Key resources

- **RAG Query** — Retrieve ranked transcript passages for a meeting and query

## AI readiness

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.

- **Score:** 34 / 100
- **Maturity:** Non-Ready
- **Dimensions:**
  - Foundational Compliance: 100 / 100
  - Developer Experience & Jentic Compatibility: 69 / 100
  - AI-Readiness & Agent Experience: 12 / 100
  - Agent Usability: 94 / 100
  - Security: 100 / 100
  - AI Discoverability: 50 / 100
- **View full report:** https://jentic.com/apis/bindstack.ai/bindstack/scorecard
- **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/
- **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview

### Score it yourself

Every API in the directory is allowlisted, so you can re-score it with no key required.

- **Score your own API:** https://jentic.com/scorecard.md
- **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md

```sh
npx @jentic/api-scorecard-cli score <openapi-url>
```

## Why Jentic

- **Setup:** The Bindstack RAG query endpoint declares no credential in its spec, so there is nothing to authenticate by hand. Through Jentic you install once, import Bindstack from the API Directory, and your agent calls the query endpoint with a meeting_code and query.
- **Permission scoping:** Bindstack carries the meeting_code in the request body, so a rule bounds which operations your agent may call rather than which meeting. With one query operation on the surface, you allow that single read operation and nothing else.
- **Credential handling:** This endpoint requires no credential in its OpenAPI spec, so there is nothing to hold. If Bindstack later adds authentication, your own Jentic One instance keeps the secret encrypted and injects it at execution time, never in the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'search a meeting transcript', and Jentic returns the Bindstack query operation with its input schema so the agent calls it with the right meeting_code and query.

## Related APIs

- **Pinecone** — Pinecone is a vector database for building your own semantic search and retrieval
- **Cohere** — Cohere provides embeddings and rerank endpoints for custom retrieval pipelines
- **OpenAI** — OpenAI models compose answers from the passages Bindstack retrieves

## FAQ

### Is there a Bindstack MCP server?

You don't need an MCP server to give your agent Bindstack. Jentic connects it directly from the API Directory: import it and your agent calls the retrieval-augmented query operation. That keeps your agent's context free of an extra server's tool definitions.

### What authentication does the Bindstack RAG Query API use?

The query endpoint does not declare any authentication in its OpenAPI spec. If your deployment requires a key, Jentic keeps it encrypted in your own instance and injects it at call time, so it never reaches the agent's context.

### Can I search a meeting transcript with the Bindstack RAG Query API?

Yes. POST `/v1/external/rag/query` takes a meeting_code and a query and returns the ranked results array with each passage's similarity score, plus the meeting_title and total_results for the match.

### How do I query a meeting transcript through Jentic?

Search Jentic for 'query a meeting transcript', load the input schema, and execute POST `/v1/external/rag/query` with a meeting_code and query. 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 Bindstack RAG Query API?

Yes. Bindstack carries the meeting_code in the request body, so a rule bounds which operations your agent may call rather than which meeting. With a single query operation, you allow that one read operation and nothing else, and every call it makes is logged.