Search past Claude Code conversations. Supports filter syntax: project:name, before:date, after:date, tool:name. Dates can be relative (7d, 30d, 1w) or ISO (2024-01-15).
AI agents call search_history to retrieve information from ClaudeHistoryMCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs read-only retrieval of conversation history data. It searches, filters, and returns information about past sessions with no capability to modify, delete, or trigger external actions. The filter syntax (project, before, after, tool) enables targeted queries but does not change any underlying data.
From the tool's definition Tool description explicitly states it 'Search[es] past Claude Code conversations' with filter syntax for querying historical data.
Documented attack patterns abuse exactly the kind of access search_history gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and ClaudeHistoryMCP, and nothing reaches the server without passing your rules. This is the rule we recommend for search_history:
{
"version": "1",
"default": "deny",
"tools": {
"search_history": {}
}
}search_history is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Search past Claude Code conversations. Supports filter syntax: project:name, before:date, after:date, tool:name. Dates can be relative (7d, 30d, 1w) or ISO (2024-01-15). It is categorised as a Read tool in the ClaudeHistoryMCP MCP Server, which means it retrieves data without modifying state.
Register the ClaudeHistory MCP server in PolicyLayer and add a rule for search_history: 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 ClaudeHistoryMCP. Nothing to install.
search_history 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 search_history 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 search_history. 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.
search_history is provided by the ClaudeHistory MCP server (jhammant/claudehistorymcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from ClaudeHistoryMCP, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
9 ClaudeHistoryMCP tools catalogued and risk-classified — across an index of 43,000+ MCP servers.