Keep the creative work in the conversation
Connect Peelaway, describe the image change to your AI assistant, and bring the finished edit back into the same conversation.
One workflow from request to result
Use MCP for tool calling or the REST contract for direct integration. Both paths share the same jobs, credits, and result model.
MCP tools
Use edit_image, get_job_status, and get_job_result from a compatible MCP client.
REST fallback
Call the OpenAPI-documented REST endpoints when direct HTTP fits the product or runtime better.
Durable job states
Each request returns an ID and moves through pending, done, or error so the caller can handle terminal outcomes explicitly.
Signed result delivery
Fetch a short-lived signed result URL, or provide a public HTTPS webhook URL for a signed terminal callback.
Machine-readable discovery
Inspect llms.txt, SKILL.md, OpenAPI 3.1, and the MCP manifest before calling the service.
Batch submissions
Submit image collections within the account plan's documented batch and concurrency limits.
Agent access without a separate product
A thin tool wrapper can submit an edit, but production automation also needs authentication, idempotency, status, failure handling, and a retrievable result.
Peelaway exposes those controls through MCP and REST. Your agent can submit a supported image edit, inspect the job state, and retrieve a signed result without introducing a second image pipeline.
Compatible MCP clients discover Peelaway sign-in from the endpoint. A revocable pk_ key remains available when a client does not support MCP OAuth.
MCP and REST use the same asynchronous jobs, status values, credits, and result delivery behavior.
REST callers can use an Idempotency-Key so a retried request does not create a duplicate charge.
The result includes dimensions, format, timing, and credit metadata. The calling workflow should still validate image quality for its use case.
Frequently asked questions
Use the contract from your existing stack
Runtimes that can send authenticated HTTP requests can call the REST API. MCP-compatible clients can use the MCP endpoint when their transport and authentication options match Peelaway's setup. The examples below are possible environments, not claims of native or tested integrations.
Coding agents and AI editors
Agent frameworks
MCP-compatible clients
Model providers and platforms
Automation and workflow tools
Discover, submit, and receive the result
Discover the contract, submit once, and let a signed webhook deliver the result.
# LLM-readable overview
curl https://api.peelaway.io/llms.txt
# Structured skill definition
curl https://api.peelaway.io/SKILL.md
# Machine-readable spec
curl https://api.peelaway.io/openapi.json# Submit
POST /api/process
{"image":"<base64>","prompt":"remove the sign","webhook_url":"https://you.app/hook"}
# → {"job_id": "abc123"}
# Signed callback when complete
POST https://you.app/hook
# → {"status": "done",
# "result_url": "https://.../signed"}
# Polling fallback
GET /api/process/status?job_id=8f4a17000e17a1b2c3d4e5f607182930
# → {"status":"done","processing_time_ms":8420}
GET /api/process/result?job_id=8f4a17000e17a1b2c3d4e5f607182930
# → {"url":"https://.../signed-url"}
Let your next idea take shape
Choose a focused creative tool, start with a photo, and shape it into a polished image. Start with free credits.