Strike48 Labs Strike48 Labs

Prospector Studio

Prospector Studio is Strike48's AI-powered platform for building and managing autonomous agents, workflows, and knowledge bases. It serves as the backend orchestration layer that all Strike48 connector apps integrate with.

Prospector Studio

What is Prospector Studio?

Prospector Studio provides a web-based interface for:

  • Conversational AI — Chat with configurable agents backed by LLM providers (Claude and more)
  • Agent Management — Create agents with custom tools, integrations, and knowledge bases
  • Workflow Automation — Build multi-step workflows with conditional routing, approval gates, and artifact management
  • Knowledge Bases — Upload documents and enable RAG (Retrieval-Augmented Generation) for context-aware responses
  • Feature Flags — Control feature rollouts at runtime

Key Features

Chat Interface

Interactive conversations with AI agents. Each conversation maintains full context and can leverage tools, knowledge bases, and workflow triggers.

Agents

Autonomous agents that combine LLM capabilities with custom tools and skills. Agents can execute actions, query knowledge bases, and trigger workflows on behalf of users.

Knowledge Bases

Document-backed RAG powered by pgvector. Upload PDFs, DOCX, CSV, JSON, and other formats — Prospector Studio automatically chunks, embeds, and indexes content for semantic search.

Workflows

Visual workflow builder with support for:

  • Container, Python, SQL, and LLM task types
  • Conditional branching and approval gates
  • Artifact management between steps
  • Real-time execution monitoring

MCP Servers

Connect external tools, resources, and prompts to agents via the Model Context Protocol. Browse available servers, configure connections, and attach tools to specific agents.

Connectors

External services that integrate with Prospector Studio over a persistent gRPC bidirectional stream. Connectors provide tools, data pipelines, and even full web UIs to the platform.

Next Steps