AI agents call export_to_prompt to retrieve information from Spec Kit without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Exporting a specification to a prompt format is a read/transformation operation. It retrieves an existing specification and converts it to a different format (Claude Code prompt). There are no side effects, deletions, or financial implications. The sibling tools (create_spec, get_spec, list_specs) confirm this server manages specs, and export is a read-like operation.
From the tool's definition "Export specification to Claude Code prompt format" — this exports/reads existing spec data into a formatted output; no mutation of data is implied.
Attacks that exploit this kind of access
Export specification to Claude Code prompt format. It is categorised as a Read tool in the Spec Kit MCP Server, which means it retrieves data without modifying state.
Register the Spec Kit MCP server in PolicyLayer and add a rule for export_to_prompt: 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 Spec Kit. Nothing to install.
export_to_prompt 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 export_to_prompt 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 export_to_prompt. 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.
export_to_prompt is provided by the Spec Kit MCP server (@fast-kit/spec-kit). 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 →