Skip to content

Latest commit

 

History

History
173 lines (128 loc) · 5.95 KB

File metadata and controls

173 lines (128 loc) · 5.95 KB

Mindee PHP SDK

Use this skill for Mindee V2 integrations with the official PHP SDK.

Scope

  • Use the official mindee/mindee PHP SDK.
  • Focus on SDK-based integration patterns only.
  • Do not suggest direct HTTP calls, cURL, or non-SDK integrations.
  • Do not use undocumented SDK internals.

Primary documentation

SDK overview

Client setup

Model parameters

Load local files

Load remote URLs

Send files and URLs

Process responses

Handle errors

Handling responses by model type

Extraction

Split

Crop

Classification

OCR

Default workflow

When answering questions, follow this order:

  1. Initialize the SDK client.
  2. Configure $modelId and other inference parameters.
  3. Load the input source.
  4. Optionally adjust the file before upload.
  5. Send with polling or webhooks.
  6. Process the response.
  7. Handle errors and retries.

Answering rules

  • Base answers on the documentation above.
  • Prefer documented SDK methods and patterns.
  • Use environment variables for API keys in production.
  • Reuse a client instance when possible.
  • Prefer polling for simple examples.
  • Prefer webhooks for production or high-volume workflows.
  • If a feature is not documented, say it is not officially supported.
  • If a user asks for code, keep examples minimal and working.

Code sample rules

  • Use PHP examples only.
  • Use the official mindee/mindee Composer package.
  • Show use imports explicitly.
  • Include the exact documented class and method names.
  • Use placeholders like MY_API_KEY, MY_MODEL_ID, and /path/to/file.pdf.
  • Keep samples focused on one task.

Preferred example topics

Client initialization

Use:

  • new \Mindee\V2\Client($apiKey)
  • new \Mindee\V2\Client() with MINDEE_V2_API_KEY environment variable

Input loading

Use:

  • PathInput — load from a local file path
  • BytesInput — load from raw binary data (provide filename)
  • Base64Input — load from a base64-encoded string (provide filename)
  • FileInput — load from a PHP file resource/handle
  • UrlInputSource — load from an HTTPS URL

Sending documents

Use:

  • $client->enqueueAndGetResult(ResponseClass::class, $inputSource, $params) for polling
  • $client->enqueue($inputSource, $params) for webhooks

Response handling

Use:

  • $response->inference
  • $response->getRawHttp()
  • LocalResponse for webhook payloads
  • HMAC signature validation via $localResponse->isValidHMACSignature() when relevant

File preparation

Use:

  • $inputSource->pageCount — property, requires Ghostscript
  • $inputSource->compress(...) — compress before upload, requires Ghostscript
  • $inputSource->applyPageOptions(...) — trim or remove pages
  • $inputSource->fixPdf()

Avoid

  • Direct REST examples
  • cURL examples
  • Manual authentication header construction
  • Bearer token examples for API keys
  • Non-PHP examples
  • V1 examples unless the user explicitly asks for V1

If the user is unclear

Ask for only what is needed:

  • input type: local file or URL
  • delivery pattern: polling or webhook
  • model ID
  • runtime context: web app, CLI, or background worker

Output style

  • Be concise.
  • Answer with runnable examples when code is requested.
  • Link to the most relevant doc section.
  • Do not overwhelm the user with every option.
  • Start with the documented default path.

Agent Instructions: Querying The Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the documentation URL with the ask query parameter. Include php+sdk+-+ at the beginning of the question to get answers specific to this library:

GET https://docs.mindee.com/integrations.md?ask=php+sdk+-+<question>

The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.