// MCP TOOL REFERENCE

KATASHIRO TOOLS

9 tools from the Katashiro MCP Server, categorised by risk level.

READ TOOLS

analyze_content Analyze text content for keywords, complexity, and sentiment extract_entities Extract entities (persons, organizations, locations, dates, URLs, emails) from text knowledge_query Query the knowledge graph for nodes and relationships web_scrape Scrape content from a URL web_search Search the web for information using DuckDuckGo or SearXNG

WRITE TOOLS

generate_report Generate a structured report from sections generate_summary Generate a summary of the provided content knowledge_add_node Add a node to the knowledge graph

EXECUTE TOOLS

deep_research Execute iterative deep research on a topic with automatic convergence detection, gap analysis, and knowledge synthesis

// FAQ

How many tools does the Katashiro MCP server have? +

The Katashiro MCP server exposes 9 tools across 3 categories: Read, Write, Execute.

How do I enforce policies on Katashiro tools? +

Use Intercept, the open-source MCP proxy. Write YAML rules for each tool — rate limits, argument validation, or deny rules — then run Intercept in front of the Katashiro server.

What risk categories do Katashiro tools fall into? +

Katashiro tools are categorised as Read (5), Write (3), Execute (1). Each category has a recommended default policy.

Let agents act without letting them run wild.

Deterministic policy on every MCP tool call. Per-identity grants. Full audit log.

GET STARTED →