canonical: https://jentic.com/apis/app.br/app

# Br Meta Agent Tools

The Meta Agent Tools API indexes Model Context Protocol servers, agent skills, and plugins in a public, searchable directory and runs security reviews of remote servers before an agent installs them. You can browse and search listings by kind, provenance, and category, register your own server or skill for moderation, and read a listing's measured security level, announced tools, and README. It also exposes feeds, index signals from external rating sources, and a private monitor that tracks the tools a given agent project has installed and the inventory changes they report.

## For AI agents

Search a directory of MCP servers, agent skills, and plugins, read their security reviews, and register or monitor listings. Authenticates with a bearer token, an account API key, or an agent wallet credential; many browse endpoints are public.

## Scope

Does not host or execute the listed servers, run their tools, or manage your agent runtime. Use for discovering, reviewing, and monitoring agent tool listings only.

## Capabilities

- Search the public directory of Model Context Protocol servers, agent skills, and plugins
- Register a server, skill, or plugin as a listing that enters moderation as pending
- Read a remote server's security review, measured signals, and change history
- Check a server address before installing it using signals already collected
- Watch a remote server or skill and track its current review and watch slots
- Monitor a private agent project's installed inventory and committed changes

## Use cases

### Agent-Vetted Tool Discovery

An AI agent connected through Jentic searches the directory for a server that matches a task, reads its security review and announced tools, and only then decides whether to install it. Because the directory returns a measured security level and the signals behind it, the agent can screen out risky servers instead of installing whatever it finds first.

Example prompt: Search the directory for a server that sends Slack messages, read its security review, and report whether its level is safe to install

### Publishing a Tool Listing

A maintainer registers a Model Context Protocol server, skill, or plugin as a listing so other developers and agents can discover it. The listing enters as pending for moderation, and editing its public content sends it back for review, so the directory stays curated rather than open spam.

Example prompt: Register a new Model Context Protocol server listing with its repository URL and category so it enters moderation

### Monitoring Installed Agent Tools

A team runs a private monitor project that records which servers and skills an agent has installed and the inventory changes each installation reports, then links those items to the directory watches that review them. This gives an operator one place to see what an agent is running and whether any of it has changed security posture.

Example prompt: List the installed inventory for a monitoring project and flag any item whose linked security review has changed

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| GET | `/api/listings` | Search the public listings |
| POST | `/api/listings` | Register a listing for moderation |
| GET | `/api/listings/{id}` | Get one listing |
| GET | `/api/listings/{id}/security` | Get a listing's security review |
| GET | `/api/check` | Check a server address before installing |
| GET | `/api/facets` | Get directory counts by kind, provenance, and category |

## Key resources

- **Listings** — Directory entries for servers, skills, and plugins, with search, facets, comments, and likes
- **Security reviews** — Measured security level, signals, and change history for a remote server or skill
- **Watches** — Followed servers and skills, their current review, slots, and offers
- **Monitor projects** — Private projects tracking an agent's installed inventory and committed changes

## 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:** 56 / 100
- **Maturity:** Foundational
- **Dimensions:**
  - Foundational Compliance: 41 / 100
  - Developer Experience & Jentic Compatibility: 63 / 100
  - AI-Readiness & Agent Experience: 57 / 100
  - Agent Usability: 69 / 100
  - Security: 58 / 100
  - AI Discoverability: 49 / 100
- **View full report:** https://jentic.com/apis/app.br/app/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:** Wiring the Meta Agent Tools API by hand means juggling several token types (a guest token, an account API key, and an agent wallet credential), sorting public endpoints from authenticated ones, and signing an x402 payment header for metered calls. Through Jentic you install once, import it from the API Directory, store your token once, and your agent calls it.
- **Permission scoping:** You choose which operations your agent may call, so you can allow browsing listings and reading security reviews while leaving listing creation, comments, and watch purchases out. The listing and monitoring-project ids travel in the URL path, so a rule can also bound the agent to one listing or one project.
- **Credential handling:** Your Meta Agent Tools token 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.
- **Discovery method:** Agents search Jentic by intent such as 'find a tool server for email' or 'read a server's security review', and Jentic returns the matching Meta Agent Tools operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

## Related APIs

- **Snyk API** — Scans code, dependencies, and containers for known vulnerabilities and license issues
- **Stoplight API** — Designs, documents, and governs OpenAPI definitions for internal and public APIs
- **ReadMe API** — Hosts developer hubs and interactive API reference documentation

## FAQ

### Is there a Meta Agent Tools MCP server?

Yes. Meta Agent Tools exposes its own Model Context Protocol endpoint. You can also connect it directly through Jentic: your agent gets the directory's full 146-endpoint surface without loading another server's tool definitions into its context, your credential is injected at call time by your self-hosted instance instead of living in a config file, and your rules decide which operations the agent may call.

### Can I limit what my agent is allowed to do with the Meta Agent Tools API?

Yes. Write a rule that allows only the listing-search and security-review read operations, so the agent can discover and vet tools but cannot create listings, post comments, or buy watch slots, and every call it makes is logged. Because listing and monitoring-project ids sit in the URL path, you can also pin the agent to a single listing or project.

### What authentication does the Meta Agent Tools API use?

The Meta Agent Tools API authenticates with bearer tokens (a guest token, an account API key, or an agent wallet credential) plus an API key header for x402 per-call payments, per its OpenAPI spec. Many browse and read endpoints are public and need no token. Through Jentic the token is stored encrypted by your own instance and added at call time, so it never appears in the agent's prompt or logs.

### Can I check a server's security before my agent installs it?

Yes. The security review for a listing returns the remote server's measured level, the signals behind it, the tools it announced at its last check, and any changes since. A separate pre-install check reports what the directory already knows about a server address without visiting it, so an agent can screen a tool before adding it.

### What are the rate limits for the Meta Agent Tools API?

The OpenAPI spec does not publish a global rate limit for the Meta Agent Tools API. It does cap comment writes at 20 per hour per owner, and watch capacity is governed by your plan; for other limits, see the Agentalog documentation at https://agentalog.com.

### How do I discover and vet agent tools with the Meta Agent Tools API through Jentic?

Import the Meta Agent Tools API from the Jentic API Directory, then have your agent issue a request such as 'search the directory for a GitHub tool server and read its security review'. Jentic matches the intent to the search and security-review operations and returns their input schemas so the agent builds the right calls, with your stored token injected at call time. To run it on your own infrastructure, install Jentic One from its GitHub repo.