Logo SVG copied to clipboard
Backed by •

The API Resilience Layer

Manifest prevents apps and agents from breaking because of failed API calls.

Start for free
Where it hurts

Why you keep having API errors in prod

  • APIs change suddenlyBreaking changes ship in every API you depend on, internal or external. The warning email never reaches the person who owns the call.
  • Requests built by agentsAn agent composes the call from what it believes the API accepts. When it is wrong, nothing in your codebase can catch it.
  • Runtime-only errorsEmpty fields, wrong formats, out-of-range values. They only exist in production, after every check went green.
Why it’s broken

Manifest fills everyone else’s blind spot: your product’s connections

Manifest is the missing piece that ensures the reliability of connected products.

Code reviewers don’t check APIs

Code review tools strictly focus on your codebase’s code quality, not its dependencies.

Observability tools don’t act

Manifest reduces downtime by fixing issues on the fly with verified patches.

iPaaS only works with popular services

Manifest works with all APIs: internal and external, popular or not.

How it works

Forget about API errors

Manifest is the resilience layer between your code and every API it calls. A failed request is repaired from a knowledge base of proven fixes and retried, in real time.

1. Add Manifest to your app or agent

3 lines of code, available for Node.js, Python and PHP, with plugins for n8n and for agents like Hermes and Claude Code.

index.js
import { manifest } from 'manifest';

manifest(); // Once, at startup.
// Keep making your API calls as usual.

2. Cartography of what you depend on

From the calls it sees, Manifest draws the map of every external service and MCP server your product talks to. The bigger and the closer a service, the more your product leans on it.

App Cal.com 132K req 326/561 healed Hacker News 41K req 88/149 healed DeepSeek 204K req 668/1.1K healed K Knowledge Management 68K req 131/228 healed OpenAI 912K req 3.3K/5.8K healed HubSpot 156K req 402/694 healed Asana 93K req 245/421 healed C Custom CRM 318K req 742/1.3K healed Supabase 187K req 509/878 healed

3. Make your app reliable

Every service carries its own record: what it costs you in latency, what share of your calls it rejects, and the change coming to it. Manifest names the fix to make at the source, for your engineer or your coding agent.

4. Manifest heals the failed request

The error is matched against a knowledge base of reviewed, deterministic patches, then the request is repaired and retried. Even if you have never seen this error: a fix learned on any Manifest customer’s traffic applies to yours.

Who it is for

Built for products that depend on other people’s APIs

Manifest lives in the code that calls APIs you do not control.

Products built on integrations

Unified APIs, connector platforms, e-commerce and logistics SaaS. Every connector breaks one day. Your customers won’t notice.

Agents that call APIs and MCP tools

Your agent builds requests at runtime. Manifest makes sure they land.

Internal and legacy APIs

Poorly versioned, still in production. Keep the callers running while the owners fix them.

Workflow automation

One failed request used to fail the whole run. Now the step heals and the workflow finishes.

Real-time self-healing

We turn 4xx API errors into 2xx successes in real time

So your apps, workflows and agents stay up.

Without Manifest

  1. 1
    ResponseHTTP 422

    HTTP 422

    {
  "param": "scheduled_at",
  "message": "Input should be a valid
     datetime, expected
     2026-09-08T18:30Z"
}

With Manifest

  1. 1
    ResponseHTTP 422

    HTTP 422

    {
  "param": "scheduled_at",
  "message": "Input should be a valid
     datetime, expected
     2026-09-08T18:30Z"
}
  2. 2
    Patch appliedfrom the knowledge base
    "template": "push-day",
"scheduled_at": "08/09/2026 18:30"
"scheduled_at": "2026-09-08T18:30Z"
  3. 3
    RetryHealedHTTP 200

    HTTP 200

    {
  "error": null,
  "statusCode": 200
}

Stop letting bad API requests take your app down.

Start for free

Errors no code review can catch

Manifest fixes errors that cannot be caught at build time.

API Breaking Change

API changed suddenly

The vendor sent a warning email a few months ago to someone on the team, but it got lost and no one read it.

POST /v1/orders
  Api-Version: 2026-07-29
  {
    "shippingMethod": "express"
    "shipping_method": "express"
  }
Non-deterministic call

An agent asks for a date the API rejects

An agent built the request at runtime and asked for one day more than the endpoint allowed.

POST /invoices/export
  {
    "account_id": "acc_2c7f",
    "from": "2026-08-01",
    "to": "2026-09-01",
    "to": "2026-08-31",
    "format": "csv"
  }
Codebase error

Request sent before async event

Resulting in an empty field sent to the API. Not caught by any code reviewer.

POST /coach/messages
  {
    "thread_id": "thr_8f2a",
    "user_id": "",
    "user_id": "Yzcc…HnL2",
    "text": "Done for today",
    "sent_at": 1788633727
  }
Pricing

Start free and scale when you need to

Starter

Free

Implement self-healing now

Start for free

No credit card required

What’s included

  • API request self-healing
  • Access to your self-healing data

Enterprise

Custom pricing

For teams with larger or more specific needs.

What’s included

  • Self-healing at scale
  • Access to your self-healing data
  • On-prem deployment
  • Custom rules
  • User management and SSO
  • Dedicated Slack support
  • Custom invoicing and payment terms

Prevent downtime due to bad API requests

Don’t wait until your production breaks. Start with Manifest now and turn failed requests into successful ones.

Start for free