Api Webhooks logo

Api Webhooks

Organization
rejot-dev
api-webhooks

Configure inbound API webhooks end to end. Use when the user needs a webhook/callback URL, provides webhook auth or sample payloads, defines provider events, or routes webhook.received deliveries into cataloged automation events.

Overview

Publisherrejot-dev
Repositoryfragno
Skill nameapi-webhooks
Stars
62
Forks
6
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by rejot-dev on GitHub. Read the source before you install it.

Installation

Install the Api Webhooks AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/rejot-dev/fragno.git /tmp/fragno
mkdir -p .claude/skills
cp -r /tmp/fragno/apps/backoffice/content/static/skills/api-webhooks .claude/skills/api-webhooks
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Api Webhooks in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Api Webhooks on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Api Webhooks is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

API Webhooks

Configure one coherent webhook contract: endpoint, event source, cataloged provider events, routes, and verification. Use codemode and return a final inventory so the user can see exactly what was registered.

Setup sequence

1. Reserve the endpoint URL

Immediately create a draft endpoint with a stable lowercase endpointId derived from the provider name, then give the user the exact publicUrl returned by the tool.

js
const draft = await api.createWebhookEndpoint({
  endpointId: "example-provider",
  name: "Example Provider",
  status: "draft",
  verification: { type: "none" },
  deliveryIdentity: { type: "jsonBodyPath", path: ["webhookId"] },
  auth: { type: "none" },
});

return { webhookUrl: draft.publicUrl };

Creating the endpoint also provisions an event source named after its endpointId. For example, endpoint example-provider owns source example-provider. Source provisioning is asynchronous. If an immediate events.catalogCreate reports that this source does not exist, retry that operation a bounded number of times for this specific error. Keep the endpoint and source identifiers unchanged.

Draft endpoints reserve a stable URL and reject deliveries with WEBHOOK_ENDPOINT_DRAFT and "Webhook endpoint is not configured yet" until activated.

2. Establish the webhook contract

Collect the information required to make deliveries deterministic:

  • example JSON payloads for every event type the user wants to receive;
  • relevant delivery, signature, and discriminator headers;
  • a provider delivery ID from a header or payload path;
  • authentication details;
  • endpoint verification challenge details, when applicable;
  • the payload field that discriminates event types, such as type.

Keep the endpoint in draft while required details are missing. The endpointId is the only value to deduce without evidence.

Choose deliveryIdentity from a stable provider identifier such as webhookId, event.id, id, or a provider delivery-ID header.

Authentication choices are none, bearer, apiKey, basic, and hmac. Treat “webhook secret,” “signing secret,” “key,” or a secret-shaped webhook value as an HMAC clue and collect the algorithm, signature location and name, encoding, optional prefix, and signed-payload format. Use bearer authentication only when the provider explicitly sends Authorization: Bearer ....

Setup guardrails

  • The API capability is available automatically for the current scope. Webhook endpoints share its storage and public URL configuration.
  • Use the shared api capability directly; connections.setup is for outbound API connections.
  • Give the user the complete, fully qualified receiving URL returned as webhook.publicUrl. Refer to it as the webhook URL, not the public base URL.
  • Assemble the final URL only from the tool result. The documented URL shape is explanatory, not a substitute for webhook.publicUrl.
  • Ask for example payloads and relevant headers before choosing delivery identities, schemas, or routes.
  • Derive only the endpointId without evidence. Keep the endpoint in draft while authentication, signature, verification, payload, header, discriminator, or delivery-identity details are unknown.
  • Store authentication secrets through endpoint tools and omit secret values from summaries, schemas, event examples, and completion responses.

3. Register provider events

For each distinct provider event, call events.catalogCreate with:

  • source: the webhook endpointId;
  • a stable eventType derived from the provider event discriminator;
  • a useful label and description;
  • a payload schema based on the projected event payload;
  • a representative example;
  • enabled: true.
js
const registered = await events.catalogCreate({
  source: "example-provider",
  eventType: "record.created",
  label: "Example Provider record created",
  description: "A record was created in Example Provider.",
  payloadSchema: {
    type: "object",
    properties: {
      recordId: { type: "string" },
      data: { type: "object" },
    },
    required: ["recordId"],
    additionalProperties: true,
  },
  example: { recordId: "rec_123", data: { status: "new" } },
  enabled: true,
});

The registered schema is the destination contract. Every route projection must produce all its required fields.

4. Route accepted deliveries

Accepted deliveries first produce api.webhook.received. Create one reclassification route for each cataloged provider event.

Every route must be endpoint-scoped. When the provider carries multiple event types, also match its event discriminator. Without the endpoint matcher, a route consumes deliveries from every API webhook endpoint.

The route matcher and projection inspect the complete automation event envelope. Webhook fields are under $.payload; provider body fields are under $.payload.body.

js
const route = await router.create({
  id: "example-provider-record-created",
  name: "Classify Example Provider record-created webhooks",
  enabled: true,
  trigger: {
    kind: "event",
    source: "api",
    eventType: "webhook.received",
    matcher: {
      all: [
        { path: "$.payload.endpointId", op: "eq", value: "example-provider" },
        { path: "$.payload.body.type", op: "eq", value: "record.created" },
      ],
    },
  },
  action: {
    kind: "reclassify_event",
    source: "example-provider",
    eventType: "record.created",
    payload: {
      kind: "projection",
      fields: {
        recordId: "$.payload.body.recordId",
        data: "$.payload.body.data",
      },
    },
  },
  description:
    "Reclassifies Example Provider record.created deliveries as example-provider record.created events.",
});

Projection paths always start with $.. Compare the projection against the destination event schema before creating the route; every required destination field must resolve for a matching delivery.

5. Activate the endpoint

Update the draft only after the webhook contract is known. When the user supplies the complete contract before setup begins, creating the endpoint directly as active is also valid.

js
const webhook = await api.updateWebhookEndpoint({
  endpointId: "example-provider",
  status: "active",
  verification: { type: "none" },
  deliveryIdentity: { type: "jsonBodyPath", path: ["webhookId"] },
  auth: { type: "none" },
});

For HMAC, configure the exact provider contract:

js
const webhook = await api.updateWebhookEndpoint({
  endpointId: "example-provider",
  status: "active",
  verification: { type: "none" },
  deliveryIdentity: { type: "jsonBodyPath", path: ["webhookId"] },
  auth: {
    type: "hmac",
    secret: "whs_secret_123",
    algorithm: "sha256",
    signature: {
      location: "header",
      name: "x-provider-signature",
      encoding: "hex",
      prefix: "sha256=",
    },
    signedPayload: { type: "rawBody" },
  },
});

HMAC decision guide

  • algorithm: use the provider's documented sha1, sha256, or sha512 algorithm.
  • signature.location: use header or query according to where the provider sends the signature.
  • signature.name: preserve the exact header or query-parameter name.
  • signature.encoding: use the documented hex, base64, or base64url encoding.
  • signature.prefix: include it only when the provider prefixes the encoded signature, such as sha256= or v1=.
  • signedPayload: use rawBody when the provider signs the unmodified request body. Use timestampedBody when it signs a prefix, timestamp, delimiter, and raw body.
  • For timestampedBody, preserve the exact prefix, timestampHeader, and delimiter, and use the provider's replay-tolerance window. An empty prefix is valid only when the provider signs the timestamp first without a prefix.

Verification challenge guide

Use { type: "none" } for ordinary delivery-only endpoints. Use challenge when the provider sends a synchronous value that must be echoed before deliveries begin.

A challenge configuration specifies:

  • the exact HTTP method;
  • a present predicate when the source only needs to exist, or an equals predicate when it must match a fixed value;
  • a match source from a header, query parameter, or JSON body path;
  • the header, query parameter, or JSON body path whose value the response echoes.

Slack-style signed challenge example

js
const webhook = await api.createWebhookEndpoint({
  endpointId: "slack",
  name: "Slack",
  status: "active",
  verification: {
    type: "challenge",
    method: "POST",
    when: {
      type: "equals",
      source: { type: "jsonBodyPath", path: ["type"] },
      value: "url_verification",
    },
    response: {
      type: "echoText",
      source: { type: "jsonBodyPath", path: ["challenge"] },
    },
  },
  deliveryIdentity: { type: "jsonBodyPath", path: ["event_id"] },
  auth: {
    type: "hmac",
    secret: "signing-secret",
    algorithm: "sha256",
    signature: {
      location: "header",
      name: "x-slack-signature",
      encoding: "hex",
      prefix: "v0=",
    },
    signedPayload: {
      type: "timestampedBody",
      prefix: "v0:",
      timestampHeader: "x-slack-request-timestamp",
      delimiter: ":",
      toleranceSeconds: 300,
    },
  },
});

return webhook.publicUrl;

6. Verify end-to-end

Tell the user when the endpoint is ready for a provider test delivery. Send a representative delivery when possible, then inspect both queues:

js
const [apiHooks, automationHooks] = await Promise.all([
  hooks.list({ fragment: "api", pageSize: 20 }),
  hooks.list({ fragment: "automations", pageSize: 50 }),
]);

Verification is complete only when:

  1. the endpoint returns an accepted response;
  2. the matching onWebhookReceived API hook completes;
  3. the original api.webhook.received automation event completes;
  4. each expected reclassified <endpointId>.<eventType> event completes without projection or schema errors.

A 202 Accepted response or a pending automation hook is intermediate evidence, not completion. When a hook fails, inspect its exact projection or schema error, repair the route, send a new unique delivery, and verify again.

Completion response

Finish every webhook setup with one inventory containing:

  • endpoint name, status, and exact fully qualified publicUrl;
  • event source (endpointId);
  • every registered source.eventType;
  • every route as trigger → destination, including endpoint and discriminator matchers;
  • authentication and verification mode, without secret values;
  • end-to-end verification status.

If the endpoint remains a draft, list the missing inputs required for activation. If no provider events were registered, state that explicitly.

Event envelope reference

api.webhook.received fires after an active endpoint accepts and authenticates a delivery. Its payload contains:

  • endpointId
  • deliveryId
  • hookId
  • receivedAt
  • redacted headers
  • redacted query
  • parsed JSON body
  • contentType

Use hookId or the automation event ID for idempotency. Duplicate provider deliveries with the same endpoint and delivery ID produce the same hook ID.

Tool reference

Use codemode first.

Webhook endpoint methods:

  • api.listWebhookEndpoints lists configured endpoints.
  • api.getWebhookEndpoint inspects one endpoint.
  • api.createWebhookEndpoint creates or replaces an endpoint.
  • api.updateWebhookEndpoint updates an endpoint.
  • api.deleteWebhookEndpoint deletes an endpoint.

A direct active endpoint is appropriate when the complete contract is already known:

js
const webhook = await api.createWebhookEndpoint({
  endpointId: "example-provider",
  name: "Example Provider",
  status: "active",
  verification: { type: "none" },
  deliveryIdentity: { type: "jsonBodyPath", path: ["webhookId"] },
  auth: { type: "none" },
});

return webhook.publicUrl;

Inspect existing configuration before replacing it:

js
const endpoints = await api.listWebhookEndpoints();
const endpoint = await api.getWebhookEndpoint({ endpointId: "example-provider" });

Event and routing methods:

  • events.catalogList
  • events.catalogGet
  • events.catalogCreate
  • router.list
  • router.get
  • router.create
  • router.update
  • router.delete

Webhook endpoints use the shared api capability automatically. Return the exact tool-provided publicUrl; the receive URL shape is /api/http/:scope/webhooks/endpoints/:endpointId/events.

Frequently asked questions

What does the Api Webhooks AI skill do?

Configure inbound API webhooks end to end. Use when the user needs a webhook/callback URL, provides webhook auth or sample payloads, defines provider events, or routes webhook.received deliveries into cataloged automation events.

Why use Api Webhooks on TypingMind?

Because you install it once and use it with any model. Api Webhooks is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Api Webhooks in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/rejot-dev/fragno/tree/main/apps/backoffice/content/static/skills/api-webhooks. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Api Webhooks?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Api Webhooks?

As many as you like. As long as a model supports skills, you can use Api Webhooks with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Api Webhooks AI skill free?

Yes. It is published on GitHub by rejot-dev under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇