AI agents call ksef_extract_invoice_source to retrieve information from KSeF MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs local LLM-based extraction and analysis of invoice fields from source documents. It retrieves and processes data but does not create, modify, delete, or execute external operations. The emphasis on non-mutation and proposal-only semantics places it firmly in the Read category with low severity, as it poses minimal risk even if misused by an AI agent.
From the tool's definition Tool description states it 'propose[s] source-backed invoice fields without mutating drafts' — the explicit 'without mutating' qualifier and use of 'propose' indicate read-only analysis. No state changes, no side effects.
Attacks that exploit this kind of access
Use a local LLM to propose source-backed invoice fields without mutating drafts. It is categorised as a Read tool in the KSeF MCP MCP Server, which means it retrieves data without modifying state.
Register the KSeF MCP server in PolicyLayer and add a rule for ksef_extract_invoice_source: 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 KSeF MCP. Nothing to install.
ksef_extract_invoice_source 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 ksef_extract_invoice_source 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 ksef_extract_invoice_source. 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.
ksef_extract_invoice_source is provided by the KSeF MCP server (olegtyshcneko/ksef-mcp). 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 →