API & AI AGENTS

Your AI agents
can do the work

Connect once and an agent gets the same access you have. It does not just answer questions. It does the job.

Let's TalkStart Free Trial
Claude · connected to Tagbox.io
Find every scene the blue sneakers appear in and drop the six best stills into the launch folder in Canva.
Searched 214 videos – 14 scenes. Six stills exported to Canva › Launch.

AI AGENTS

Not a chatbot.
A colleague with access.

01

Run a content studio on it

Teams build full content studios on top: the agent picks the shots and embeds photos and videos into social posts, website pages and campaigns - automatically.

02

Enrich from your systems

Pull data in from the systems around the library, or run the images through your own pipelines and write the descriptions back.

03

Automate the busywork

Sending each guest their upload link, sharing assets with collaborators the moment they land - the processes nobody enjoys doing by hand.

A content studio, running
"Find me videos that fit this brief"
"Create a post from our latest batch of photos"
"Pull this week's five best shots"
Arriving enriched
IMG_2114.jpg→Vertex Runner · Coral · SKU VR-204
IMG_2180.jpg→Spring 26 · Cleared for paid social
clip_final2.mp4→Launch film · Pier 27 · Highlight cut
The busywork, handled
Upload links sent to every photographer
All 1-star rated files moved out of the library
Best shots of the management team, found and filed
POST /v1/search
{ "query": "rooftop at sunset",
"results": [
"rooftop_055.jpg""tags": ["roof terrace", "sunset", "Spring 26"]
"venue_hall_012.jpg""tags": ["venue", "golden hour", "Pier 27"]
"stage_wide_004.jpg""tags": ["stage", "before doors"]
], "total": 3 }

REST API

DEVELOP DIRECTLY ON THE REST API

Search, upload, tags, metadata and collections over a documented REST API.

Read the API docs

CONNECT YOUR AGENTS

Let's TalkStart Free Trial

FREQUENTLY ASKED QUESTIONS

"Which plans include it?"

API and MCP access starts on Pro.

"What is MCP, in one line?"

The standard that lets an AI agent use a tool directly - so the agent searches your library itself instead of you pasting things into it.

"Can an agent change things, or only read?"

Both - search, tag, describe, upload, organize and share.

"What stops an agent touching something it should not?"

You do, from your agent platform. Permissions there decide what an agent may do on its own and what needs your explicit approval - so routine work runs unattended and anything you'd rather see first waits for you.

"What do teams actually build on it?"

The three we see most: an agent that pulls the right shots into a campaign brief, one that writes descriptions back onto a batch, and one that assembles a client-ready collection from a plain-language request.

If you can describe it, you can find it✦If someone's in it, you can find them✦If it was said on camera, you can search it✦If you remember one detail, that's enough✦If it's in the frame, it's in the search✦If your team shot it, your team can find it✦If you can describe it, you can find it✦If someone's in it, you can find them✦If it was said on camera, you can search it✦If you remember one detail, that's enough✦If it's in the frame, it's in the search✦If your team shot it, your team can find it✦