Read the chunks immediately before and after a query_documents result, in the same document, for more surrounding context. Pass chunkIndex from the result plus exactly one of filePath (ingest_file) or source (ingest_data). Returns the target chunk (isTarget: true) and its neighbors, ascending by ...
AI agents call read_chunk_neighbors to retrieve information from Mcp Local Rag without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a retrieval-only operation that fetches contextual document chunks. It performs no writes, deletes, executes code, or triggers external systems. The data it returns is already indexed and accessible through query_documents. The worst misuse would be information disclosure or excessive context retrieval, which is low severity.
From the tool's definition Tool reads and retrieves chunk neighbors from indexed documents. Description states: 'Read the chunks immediately before and after a query_documents result' and 'Returns the target chunk and its neighbors'. No modification, deletion, or side effects on data.
Attacks that exploit this kind of access
Read the chunks immediately before and after a query_documents result, in the same document, for more surrounding context. Pass chunkIndex from the result plus exactly one of filePath (ingest_file) or source (ingest_data). Returns the target chunk (isTarget: true) and its neighbors, ascending by chunkIndex; an out-of-range chunkIndex returns []. Defaults: before=2, after=2 (max 50 each). It is categorised as a Read tool in the Mcp Local Rag MCP Server, which means it retrieves data without modifying state.
Register the Mcp Local Rag MCP server in PolicyLayer and add a rule for read_chunk_neighbors: 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 Mcp Local Rag. Nothing to install.
read_chunk_neighbors 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 read_chunk_neighbors 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 read_chunk_neighbors. 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.
read_chunk_neighbors is provided by the Mcp Local Rag MCP server (mcp-local-rag). 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 →