revision-external-api

Manages architecture components, diagrams, attributes, tags, and templates via the Revision External API. Triggers when Revision API, architecture components, C4 diagrams, or syncing architecture data with Revision are mentioned.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "revision-external-api" with this command: npx skills add revision-org/revision-skills/revision-org-revision-skills-revision-external-api

Revision External API

REST API for managing architecture documentation in Revision workspaces.

Prerequisites

Before making any API calls, the user must provide two things:

  1. Organization URL: Each organization has its own subdomain, e.g. https://acme-company.revision.app/. This is the base URL for all API requests. Always ask the user for their organization URL — there is no default.
  2. API key: A Bearer token from the workspace settings.

Authentication

All requests require a Bearer token (API key from workspace settings):

Authorization: Bearer <api-key>

Base URL

The base URL is the organization's own Revision URL. Every organization has a unique subdomain:

https://{organization}.revision.app

For example: https://acme-company.revision.app

Important: Do not use a generic URL. Always use the organization-specific subdomain provided by the user.

Resources

ResourceEndpointsDescription
ComponentsCRUD + batch upsert + filterArchitecture components (services, databases, etc.)
DiagramsCRUD + batch upsert + filterArchitecture diagrams with component instances, relations, textareas
AttributesCRUD + batch upsertCustom attribute definitions on components
TagsCRUD + batch upsertTags for categorizing diagrams
TypesRead-onlyComponent type definitions
TemplatePOSTBulk sync of components + diagrams in a single transaction

Quick Start

List components

curl -H "Authorization: Bearer $API_KEY" \
  https://acme-company.revision.app/api/external/components

Create a component

curl -X POST -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"name": "User Service", "state": "ACTIVE"}' \
  https://acme-company.revision.app/api/external/components

Create a component with a predictable ID

curl -X POST -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"id": "user-service", "name": "User Service", "state": "ACTIVE"}' \
  https://acme-company.revision.app/api/external/components

Update a component

curl -X PATCH -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"name": "User Service", "state": "ACTIVE", "desc": "Handles user auth"}' \
  https://acme-company.revision.app/api/external/components/component-id

Standard CRUD Pattern

Every resource (components, diagrams, attributes, tags) follows the same pattern:

MethodPathAction
GET/api/external/{resource}List all (with optional query filters on components, diagrams, types)
POST/api/external/{resource}Create (single item)
GET/api/external/{resource}/{id}Get by ID
PATCH/api/external/{resource}/{id}Update by ID
PATCH/api/external/{resource}/upsert-batchBatch upsert (items with id are updated, without are created)

Note: DELETE endpoints exist but are not yet implemented (return 501).

ID behavior: If you provide an id when creating, that exact ID is used (predictable). If you omit id, one is auto-generated. Providing a predictable id is useful when you need to reference the resource elsewhere (e.g. a component's id in a component instance's componentId).

Create requests accept a single object (not an array). Returns 201 on success.

Batch upsert is the most powerful pattern: send items with id to update, without id to create, all in one request.

Component Schema

{
  "id": "string",
  "name": "string",
  "state": "DRAFT | ACTIVE | ARCHIVED",
  "desc": "string | null",
  "inlineDesc": false,
  "typeId": "string | null",
  "apiContext": "string",
  "attributes": [{ "id": "string", "value": "boolean | number | string | null" }],
  "linksTo": ["diagram-id-1", "diagram-id-2"]
}
  • id: Optional on create — provide it for a predictable ID, omit for auto-generated
  • apiContext: Optional label to group related imports (defaults to current UTC timestamp if omitted)
  • linksTo: Array of diagram IDs this component links to
  • attributes: Component attribute values (reference attribute definitions by id)

Diagram Schema

{
  "id": "string",
  "name": "string",
  "state": "DRAFT | ACTIVE | ARCHIVED",
  "url": "string",
  "desc": "string | null",
  "level": "C0 | C1 | C2 | C3 | C4 | D1 | P0 | null",
  "tags": ["tag-id-1"],
  "apiContext": "string",
  "componentInstances": [],
  "relations": [],
  "textareas": []
}

Diagram Levels

LevelMeaning
C0Landscape
C1System Context
C2Container
C3Component
C4Code
D1Deployment
P0Process

Component Instances

A component is part of the architecture model — a reusable entity that can be referenced across many diagrams. A component instance is a visual placeholder on a diagram. It can optionally link to a component via componentId, but it doesn't have to — unlinked instances are just standalone placeholders with a name and type.

Two types:

Important: position, width, and height are all optional. When omitted, Revision will automatically lay out and size the instances. Prefer omitting them unless the user explicitly asks for specific positioning or sizing — auto-layout produces better results.

Non-container (default):

{
  "ref": "unique-ref",
  "componentId": "component-id | null",
  "parent": "container-ref",
  "isContainer": false,
  "placeholder": { "text": "Name", "typeId": "type-id" }
}

Container (groups other instances):

{
  "ref": "unique-ref",
  "componentId": "component-id | null",
  "isContainer": true,
  "placeholder": { "text": "Name", "typeId": "type-id" }
}
  • ref is required and must be unique within the diagram
  • componentId links the instance to a component definition (null for placeholders)
  • Non-containers can reference a parent container via the container's ref
  • Containers cannot have a parent
  • position, width, height are optional — omit them to let Revision auto-layout and auto-size

Relations

Directed edges between component instances:

{
  "fromRef": "instance-ref-1",
  "toRef": "instance-ref-2",
  "label": "string | null",
  "desc": "string | null",
  "linksTo": ["diagram-id"]
}

Textareas

Free text on diagrams:

{
  "position": { "x": 100, "y": 100 },
  "width": 300,
  "text": "string | null"
}

Attribute Schema

Custom fields on components:

{
  "id": "string",
  "name": "string",
  "desc": "string | null",
  "type": "STRING | NUMBER | BOOLEAN | LINK | USERLIST | LIST",
  "list": ["option1", "option2"],
  "forTypes": ["type-id-1"],
  "required": false,
  "apiContext": "string"
}
  • list is only valid when type is LIST (and required for LIST)
  • forTypes: Restrict attribute to specific component types

Tag Schema

{
  "id": "string",
  "name": "string",
  "desc": "string | null",
  "color": "gray | red | orange | yellow | green | blue | purple",
  "apiContext": "string"
}

Type Schema (Read-Only)

{
  "id": "string",
  "name": "string"
}

Types are managed in the Revision UI. List via GET /api/external/types. Supports optional name query parameter for filtering.

Template Sync

Bulk sync components and diagrams in a single transaction:

curl -X POST -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"components": [...], "diagrams": [...]}' \
  https://acme-company.revision.app/api/external/template

Accepts both JSON and YAML. Components and diagrams follow their standard schemas.

Filtering & Dependencies

Filtering Components

GET /api/external/components?name=...&typeId=...&tagId=...&attributeId=...&attributeValue=...&state=...

All filters optional. attributeValue requires attributeId.

Filtering Diagrams

GET /api/external/diagrams?componentId=...&tagId=...&name=...&level=...&state=...

All filters optional.

Filtering Types

GET /api/external/types?name=...

Component Dependencies

GET /api/external/components/{id}/dependencies

Returns DependencySearchResult[] — upstream and downstream direct dependencies for a component.

Workflow: Create a Diagram with Components

A typical end-to-end flow: search for duplicates, create components, then create a diagram that references them.

  1. Search for existing duplicates — MUST do this before creating anything:

    • For each component you plan to create, search by name: GET /api/external/components?name=<name>
    • For the diagram, search by name: GET /api/external/diagrams?name=<name>
    • If matches are found → ask the user whether to reuse the existing resource or create a new one
    • If the user chooses to reuse → use the existing resource's id instead of creating a new one
    • If no matches → proceed to create
    • Skip ONLY when the user explicitly says "create a new..." or "update the existing..."
  2. List types to find the right typeId values:

curl -H "Authorization: Bearer $API_KEY" \
  https://acme-company.revision.app/api/external/types
  1. Create components (skip any that the user chose to reuse from step 1):
curl -X POST -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"id": "user-service", "name": "User Service", "state": "ACTIVE"}' \
  https://acme-company.revision.app/api/external/components
curl -X POST -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"id": "user-db", "name": "User Database", "state": "ACTIVE"}' \
  https://acme-company.revision.app/api/external/components
  1. Create a diagram with component instances and a relation:
curl -X POST -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "User Service Context",
    "level": "C2",
    "state": "ACTIVE",
    "componentInstances": [
      { "ref": "us", "componentId": "user-service" },
      { "ref": "udb", "componentId": "user-db" }
    ],
    "relations": [
      { "fromRef": "us", "toRef": "udb", "label": "Reads/writes" }
    ]
  }' \
  https://acme-company.revision.app/api/external/diagrams
  1. Verify by fetching the diagram back:
curl -H "Authorization: Bearer $API_KEY" \
  https://acme-company.revision.app/api/external/diagrams/<returned-id>

For bulk operations, use the template endpoint to sync everything in a single transaction instead.

Output Summary

After every mutation (create, update, batch upsert, or template sync), print a summary that clearly separates what was created from what was updated.

Format:

Created:
  - Component "User Service" (id: user-service)
  - Diagram "System Context" (id: system-context)

Updated:
  - Component "Auth Service" (id: auth-service) — updated name, desc

Rules:

  • Use Created for POST (201) responses and batch upsert items that had no id provided
  • Use Updated for PATCH (200) responses and batch upsert items that had an id provided
  • Always include the resource type, name, and ID
  • For updates, briefly note which fields changed
  • Omit a section if it's empty (e.g. don't print "Updated:" if nothing was updated)

Error Responses

All errors return:

{ "error": "description" }
StatusMeaning
400Validation error
401Missing or invalid API key
404Resource not found
405Method not allowed
501Not implemented (e.g. DELETE endpoints)

Error Recovery

When an API call fails:

  1. 401: Confirm the API key is correct and the Authorization header uses Bearer prefix.
  2. 400: Read the error field — it describes the validation issue. Fix the request body and retry.
  3. 404: Verify the resource ID. Use the list endpoint (GET /api/external/{resource}) to confirm the resource exists.
  4. 501: DELETE is not implemented. Use PATCH to set state: "ARCHIVED" instead.

Always verify mutations by fetching the resource back after create/update to confirm the change took effect.

Full OpenAPI Spec

For the complete OpenAPI 3.0.3 specification with all request/response schemas, see OPENAPI.md.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

claw2ui

Generate interactive web pages (dashboards, charts, tables, reports) and serve them via public URL. Use this skill when the user explicitly asks for data vis...

Registry SourceRecently Updated
General

WeChat Article Summarize

Read one or more WeChat public account article links from mp.weixin.qq.com, extract cleaned full text and optional image links, summarize each article in Chi...

Registry SourceRecently Updated
General

Openfinance

Connect bank accounts to AI models using openfinance.sh

Registry SourceRecently Updated
General

---

合同审查清单AI助手 - 5类合同+3大特殊条款,风险识别与修改建议

Registry SourceRecently Updated