AI agents call lsp_completions to retrieve information from Lsp without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries a language server for code completion metadata at a given position. It retrieves and returns data without modifying code, files, or executing operations. The blast radius is minimal—worst case, an AI agent receives spurious suggestions but cannot alter the codebase or trigger external actions. Classified as Read with low severity.
From the tool's definition Tool description states it 'Get intelligent code completion suggestions' and 'Returns available methods, properties, variables, and types' — purely informational retrieval with no side effects.
Attacks that exploit this kind of access
Get intelligent code completion suggestions at a position. Returns available methods, properties, variables, and types that are valid in that context. More accurate than text-based completion. It is categorised as a Read tool in the Lsp MCP Server, which means it retrieves data without modifying state.
Register the Lsp MCP server in PolicyLayer and add a rule for lsp_completions: 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 Lsp. Nothing to install.
lsp_completions 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 lsp_completions 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 lsp_completions. 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.
lsp_completions is provided by the Lsp MCP server (lsp-mcp-server). 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 →