# Find and Run APIs and AI Models with OpenAny

Finding a suitable API or AI model often means checking several catalogs. OpenAny lets you browse one catalog, review public prices, and approve calls individually.

## Install

```bash
npx skillstore add openanysi/openany
```

## Metadata

- Status: approved
- Slug: openanysi-openany
- Skillstore revision: r1
- Version status: missing
- Tree hash: 3d0865ad07d09137a77b754c1af7a763b402463bdbf4e9cd8a614596c5d7754b
- Author: openanysi
- GitHub username: openanysi
- Repository: https://github.com/openanysi/openany-mcp/tree/700a577156c33db92f69be2f2a39d9ad7c61b21c/plugins/openany/skills/openany
- Ref: 34a07e505f0c3a61507b491b8ace4b0486f366c6
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, env\_access
- Quality score: 64
- Public page: https://skillstore.pages.dev/skills/openanysi-openany
- Manifest: https://skillstore.pages.dev/api/skills/openanysi-openany/manifest

## Capabilities

- Retrieve catalog counts for data APIs by category and AI models by modality.
- Search for APIs and AI models using a business query or a returned filter.
- Look up an entry's input schema, API version, and published price.
- Use the published version and input schema to execute an authenticated call after user approval.
- Report the result or gateway error returned by a call.

## Use Cases

- Research Data Sources: Browse company and finance API categories and find entries relevant to a research task without making a paid call.
- Review API Requirements: Check the input schema, version, and public price of an API before deciding whether to integrate it.
- Compare Model Options: Find models for a content task and compare their documented inputs and prices before approving a trial.

## Prompt Templates

### Browse the Catalog

```
Show the available data API categories and AI model modalities. Use the counts from OpenAny. Do not run a paid call.
```

### Search for a Data API

```
Find OpenAny APIs for public company profiles. Summarize what the results offer. Do not run a paid call.
```

### Compare Model Details

```
Find two available text models for summarization, if possible. Compare their API versions, required inputs, and published prices. Do not run either model.
```

### Prepare One Approved Call

```
Find an API for public company profiles and prepare a request using only a public company name. Show the API version, destination, published price, inputs, and known side effects. Wait for my approval before running it. Do not retry if execution or charging is unclear.
```

## Limitations

- The skill needs the OpenAny MCP server connection and does not use the full catalog's tool names.
- Execution requires an OpenAny API key and may spend your balance or cause side effects.
- The skill does not declare a fixed price or verify the final charge from an HTTP success response.
- Search queries and execution inputs are sent to the remote OpenAny service.

## Best Practices

- Use the catalog results for names and counts, and check the entry details for the current published price.
- Before approving a paid call, review the destination, price, exact input, and known side effects.
- Keep the API key in your client's supported secret configuration and send only the data needed for the task.

## Anti Patterns

- Pasting the API key into chat or including it in request data.
- Treating a successful tool listing or HTTP response as proof that a paid call completed or was charged.
- Retrying a call when you cannot confirm whether it ran or incurred a charge.

## Security Audit

- Audited at: 2026-11-20T14:18:33.548\+00:00
- Summary: The backticks are Markdown, and the secret reference documents authentication without asking the assistant to read or expose the key. The intended remote calls send inputs to OpenAny. Paid calls require approval, but the skill does not require showing the user the price, data to be sent, or known side effects before approval.

## Stats

- Views: 0
- Downloads: 2
- Favorites: 0
- Popularity score: 0
