Find agent skills by outcome

prompt: mechanical engineering drawing

multi-view orthographic projection

output_expectation: flat 2D floor plan similar to architectural plan drawings

suitable for real estate listings or construction documents

3D to 2D Floor Plan Converter

{

prompt: top-down 2D architectural floor plan

black and white technical drawing

negative_prompt: 3d perspective

isometric view

6. share_amount (required): Must be a non-zero integer. If missing/invalid

omit the shareholder. (`1` is usually suspicious.)

7. share_percentage (optional): Decimal percentage (e.g.

`45.0`). If missing

* If missing/invalid

return `null`.

* If only city is present

return just the city string.

* numbers/IDs only

or unclear/garbled strings.

* Prefer <street>

<city>

3. shareholder_name (required): Must be a real

identifiable person or company name. Exclude:

* legal/notarial terms (e.g.

“Notar”)

* a valid non-zero `share_amount` (integer

EUR).

address: Other street 21

Gotham

* a valid `shareholder_name`

and

address: Some street 10

Metropolis

Data

If no valid shareholders are found (or the data is too corrupted/incomplete)

return an **empty array**: `[]`.

Respond with only valid JSON (no additional text

no markdown).

PDF Shareholder Extractor

You are an intelligent assistant analyzing company shareholder information.

prompt: STRICT IDENTITY ANCHOR:\nUse ${ref_image} as a strict identity anchor for the main subject. Preserve the same person exactly: facial structure

proportions

negative_prompt: any extra text

misspelled words

BIRTH_PLACE: {City

Country}

2026 Size Neler getirecek

{