Full-text keyword search across all indexed MD files. Returns file paths, line ranges, heading context, and content snippets ordered by FTS5 relevance. More precise than Grep — matches are scoped to semantic MD blocks (headings, paragraphs, list items). For recent changes, call md_status first. F...
AI agents call md_search to retrieve information from Md Graph without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
md_search is a search/query operation that retrieves and displays information from markdown files without creating, modifying, or deleting data. It has no side effects on the knowledge graph state. The description explicitly states it 'Returns' results and advises users to 'Read the actual file for complete context,' confirming its role as a read-only lookup tool.
From the tool's definition The tool 'Full-text keyword search across all indexed MD files. Returns file paths, line ranges, heading context, and content snippets' performs read-only retrieval operations. It 'Returns' data without modifying any files or triggering external effects.
Attacks that exploit this kind of access
Full-text keyword search across all indexed MD files. Returns file paths, line ranges, heading context, and content snippets ordered by FTS5 relevance. More precise than Grep — matches are scoped to semantic MD blocks (headings, paragraphs, list items). For recent changes, call md_status first. For link relationships, use md_navigate. Snippets are truncated — always Read the actual file for complete context. It is categorised as a Read tool in the Md Graph MCP Server, which means it retrieves data without modifying state.
Register the Md Graph MCP server in PolicyLayer and add a rule for md_search: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Md Graph. Nothing to install.
md_search is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the md_search rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for md_search. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
md_search is provided by the Md Graph MCP server (lzmw/md-graph). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →