Redactor Agent Toolkit

The fastest way to add Redactor to your product.

Integrate Redactor in days, not months.

What used to take months now takes a few days. No wiring requests and responses by hand, no stitching scattered docs together, and no first client built from scratch.

Download the toolkit Read the developer reference

Free download. Needs a Redactor license and Docker to run.

Local Redactor jobLicensed instance

Prompt

Use the Redactor Agent Toolkit. Redact vehicles in small.mp4 on our licensed instance. Keep media local. Return status and output_path.

Local result

toolkit/ open
prompt: redact vehicles in small.mp4
ops: detect_vehicles → render_output
status: complete
output_path: ./out/small_redacted.mp4

1 · toolkit open2 · redaction prompt3 · local output_path

A Redactor integration used to start the slow way

The old path burned calendar before product work started. Learn the API. Build the request and response logic. Ship the first client. Catch up every release.

01 Learn the surface first
Weeks go to reading endpoints before any customer workflow runs.

02 Build every request and response by hand
The team has to piece together the shapes from scattered references.

03 Build the first client yourself
Setup tax lands before anything customer-facing ships.

04 Re-sync every release
The first build is only the start. Maintenance drag follows it.

Stop rebuilding the Redactor client every release

Speed, the current Redactor API and setup context, your agent, your instance.

01 Faster to first running job

From decision to a running API workflow in days, not months.

02 Stays current with every release

Generated from Redactor's own codebase, so it tracks the current API instead of last quarter's docs.

03 Works with your existing agent

Cursor, Claude, Codex, Copilot, Warp, Windsurf, Gemini. Not locked to one assistant.

04 Runs on your instance

Works against your licensed Redactor instance in your Docker setup. Media stays with you.

A connected package, not a prompt wrapper

One package. The files, examples, and reference your agent needs to write a real Redactor integration.

Your agent starts from real context instead of a blank page.

✓
Deployment examples - a known path to the first successful run.
✓
TypeScript definitions - shapes that are correct before the code runs.
✓
OpenAPI reference files - the API structure the client is built from.
✓
Agent-ready integration context - less time teaching the agent what Redactor is.
✓
Local setup guidance - setup steps the agent can follow as soon as the folder is open.

Example photo

From zip to first local redaction job

Unzip. Point the agent. Run the job on your instance.

Download and open the package
Unzip the folder onto disk. That is the whole setup.

Point your agent at the folder
You do not need to feed the agent one file at a time. The package already gives the agent the files it needs.

Name the job. Run it locally.
The agent writes the integration and runs it against your licensed instance.

Example photo

If you own the integration decision

Shipping redaction inside a product

You need the first Redactor workflow without parking the roadmap for a quarter.

Standing up an internal redaction path

Clear the setup tax before the architecture debate starts.

Weighing build vs embed

The toolkit makes the time cost concrete, not just the capability claim.

Building redaction is not the hard part. Starting is.

The model is only part of the cost. Client setup, payload mapping, and release sync eat the months before a customer ever sees a redacted frame. The toolkit removes that first layer of setup work so the team spends time on the product workflow instead.

Before you download

Does it need an internet connection?

Media and processing stay local. Licensing is the one outbound step. Offline license mode exists if you need fully disconnected runs.

Do I need to know the Redactor API already?

No. The toolkit gives your team and agent the structure to write the integration.

‍

What do I need to get started?

A Redactor license, Docker, and the toolkit.

‍

Where does my media go?

Nowhere outside your environment. Processing runs locally on your Redactor instance.

‍

Which languages does it support?

Python is the clearest path today. The type contract also feeds generators for other languages.

‍

Does this lock me into one AI coding tool?

No. Use the coding agent your team already runs.

‍