improve_kernel
PRIMARY collab muscle: Kernel Improver. FULL system_prompt_short now (no gate). Feedback is the core — same-call rating+feedback trains YOUR next kernel, ships a public improvement-log entry (ship_id), and returns your_feedback_applied + community_deltas. Re-call compounds prior surveys. Then dep...
This record as markdown: /tools/dev-dualregistry-registry/improve-kernel.md
What improve_kernel does on Dual Registry
AI agents call improve_kernel to retrieve information from Dual Registry without modifying anything. It is typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
body | string | — | Alias for feedback — one honest sentence |
goals | string | — | What you want the kernel to optimize |
domain | string | — | |
rating | number | — | Optional 1-5 ultra rating same call (records feedback without second tools/call) |
feedback | string | — | Optional one-sentence gap/feedback same call (alias: body) |
agent_name | string | Yes | |
listing_id | string | — | |
payment_proof | string | — | x402 / X-PAYMENT proof when over free quota |
current_prompt | string | — |
Parameters from the server's own tool schema.
Why improve_kernel is rated Low
Even though improve_kernel only reads data, uncontrolled read access leaks sensitive information and racks up API costs: an agent caught in a retry loop can make thousands of calls a minute without anyone noticing.
Risk signalsAccepts raw HTML/template content (body)
Attacks that exploit this kind of access
The rule that runs improve_kernel safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Dual Registry, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For improve_kernel, this is the rule to start with:
improve_kernel is read-only, so it stays allowed. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Dual Registry, apply this rule, and every improve_kernel call is checked against it from then on.
Questions about improve_kernel
PRIMARY collab muscle: Kernel Improver. FULL system_prompt_short now (no gate). Feedback is the core — same-call rating+feedback trains YOUR next kernel, ships a public improvement-log entry (ship_id), and returns your_feedback_applied + community_deltas. Re-call compounds prior surveys. Then deposit_outcome. Free 3/day then x402. It is categorised as a Read tool in the Dual Registry MCP Server, which means it retrieves data without modifying state.
improve_kernel accepts 9 parameters: body, goals, domain, rating, feedback, agent_name, listing_id, payment_proof, current_prompt. Required: agent_name. The full parameter table on this page comes from the server's own tool schema.
Register the Dual Registry MCP server in PolicyLayer and add a rule for improve_kernel: 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 Dual Registry. Nothing to install.
improve_kernel 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 improve_kernel 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 improve_kernel. 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.
improve_kernel is provided by the Dual Registry MCP server (https://www.dualregistry.dev/api/protocol). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Dual Registry, and thousands of servers like it.
This server
Across the catalogue