Explore automations

Fetch posts from a LinkedIn organization's feed. Looks up the organization by vanity name and retrieves its posts with pagination support.

Fetch the authenticated user's LinkedIn member profile information including name, headline, and profile picture. (Personal/member accounts only)

Retrieve comments on a specific LinkedIn post with pagination support. Returns detailed comment threads for monitoring engagement opportunities. (Works for both personal and organization posts)

Publishes posts to LinkedIn. Supports org posts with media by using a PERSON uploadOwner for registerUpload while keeping author as the organization URN. For org+media, pass uploadOwner=urn:li:person:<id>. Uses member posts endpoint for publishing.

Manage audience interactions on LinkedIn posts. Create comments, like posts, or delete comments. Primary tool for two-way engagement and community management on LinkedIn. (Works for both personal and organization posts)

Registers media uploads with LinkedIn and returns asset URNs that can be used in post content. Handles both image and video uploads through LinkedIn's asset registration API. This is a prerequisite step before publishing posts with media. (Works for both personal and organization posts)

Create a text post on LinkedIn as the authenticated member using the UGC Posts API

Create a text post on LinkedIn as the authenticated member using the UGC Posts API

Test an agent's execution with sample input to validate configuration and behavior. ## Purpose Execute an agent with test input to verify: - Agent configuration is valid - Blueprint executes without errors - System prompt produces expected behavior - Assigned tools work correctly - Response quality meets requirements ## Required Parameters - **slug**: Agent slug to test - **input**: Test input (messages array or simple text) ## Input Formats For conversational agents, provide messages array: ```json { "slug": "support-agent", "input": [ { "role": "user", "content": "Hello, I need help" } ] } ``` For simple text input: ```json { "slug": "analyzer", "input": "Analyze this data" } ``` ## Environment Dependencies Automatically inferred from agent state (do not pass manually): - **STARSHIP_ID**: Current starship identifier ## Response & Behavior - **Output**: Agent execution result with response, status, and any errors - **Timeout**: 60 seconds for agent execution - **Common failures**: Agent not found (404), invalid configuration, execution timeout, tool errors

Removes an edge from a tool's blueprint while maintaining node integrity.

Removes a node from a tool's blueprint and automatically cleans up all edges connected to it.

Adds a new edge to an existing tool's blueprint with automatic validation and conflict detection. ## Required Parameters - **toolSlug**: Slug of the tool to modify - **edge**: New edge to add with ref, from, to, optional condition/label/metadata

Adds a new node to an existing tool's blueprint with automatic validation and conflict detection. ## Purpose Incrementally build tool blueprints by adding nodes one at a time, ensuring each addition maintains blueprint integrity. ## Required Parameters - **toolSlug**: Slug of the tool to modify - **node**: New node to add with ref, type, name, description, parameters ## Node Requirements - **ref**: Unique identifier (string) - **type**: Valid node type (start, end, llm, code, http_request, conditional, transform, mutate, tool, workflow, user_prompt) - **name**: Human-readable name - **description**: Optional description - **parameters**: Node-specific parameters object. Use node-schema tool to get the exact structure for each node type. ## Workflow 1. Fetches current tool by slug 2. Validates new node structure and uniqueness 3. Adds node to tool's blueprint 4. Updates tool with modified blueprint ## Validation - Verifies tool exists and is accessible - Checks for ref uniqueness within blueprint - Validates node type against supported types - Ensures required node fields are present - Maintains blueprint structural integrity ## Response Returns updated tool with new node in blueprint, or validation errors if addition fails

Query a library of working blueprint examples for different scenarios and patterns. ## Purpose Provides real-world, tested blueprint examples showing proper structure, node usage, error handling, and advanced patterns. ## Example Categories - **basic**: Simple linear workflows - **conditional**: Branching and decision logic - **loops**: Tool calling loops and iterations - **api_integration**: HTTP requests and API workflows - **llm_patterns**: LLM usage patterns with tools - **error_handling**: Resilient workflows with error recovery - **state_management**: Complex state transformations - **nested**: Workflows with nested blueprints ## Response Format Returns complete, executable blueprints with: - Detailed node configurations - Proper edge connections - Template expression examples - Comments explaining the pattern

Query detailed schema information for Autopilot node types including parameters, validation rules, and execution behavior. ## Purpose Provides real-time, accurate schema information for all Autopilot node types, including complex nested structures, validation requirements, and execution patterns. ## Query Capabilities - **Specific Node Type**: Get complete schema for a single node type - **All Node Types**: List all available node types with descriptions - **Parameter Details**: Deep dive into parameter schemas including nested structures - **Validation Rules**: Understand validation requirements and constraints - **Execution Patterns**: Learn how nodes execute and interact ## Node Types Available start, end, llm, agent, http_request, conditional, transform, mutate, tool, workflow, code, user_prompt ## Response Format Returns comprehensive schema with: - Parameter definitions with types and constraints - Required vs optional fields - Validation rules and ranges - Template expression support - Inline operation capabilities - Execution behavior notes

Fetch instant answers, definitions, calculations, and quick facts from DuckDuckGo.

Search Wikipedia and return results via the Wikipedia REST API.

Echo back the provided input for testing purposes.