Read the failed Git patch that could not be applied to the integration worktree.
AI agents call read_failed_patch to retrieve information from Ktx without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries diagnostic data (a failed Git patch) from the system without any side effects, modifications, or execution of code. It is purely informational, helping agents understand why a patch application failed. No data is modified, deleted, or executed.
From the tool's definition Tool name contains 'read' and description states 'Read the failed Git patch' - explicitly a retrieval operation with no modification or execution capability. Returns diagnostic information about a failed patch application.
Documented attack patterns abuse exactly the kind of access read_failed_patch gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Ktx, and nothing reaches the server without passing your rules. This is the rule we recommend for read_failed_patch:
{
"version": "1",
"default": "deny",
"tools": {
"read_failed_patch": {}
}
} read_failed_patch 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.
Read the failed Git patch that could not be applied to the integration worktree. It is categorised as a Read tool in the Ktx MCP Server, which means it retrieves data without modifying state.
Register the Ktx MCP server in PolicyLayer and add a rule for read_failed_patch: 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 Ktx. Nothing to install.
read_failed_patch 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_failed_patch 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_failed_patch. 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_failed_patch is provided by the Ktx MCP server (kaelio/ktx). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Ktx, 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.
10 Ktx tools catalogued and risk-classified — across an index of 43,000+ MCP servers.