diagnose
Score how faithfully a reproduction matches an original (MAE / PSNR / exact-match % / stray px). Returns: A JSON string with the fidelity metrics.
This record as markdown: /tools/design-reprise-reprise/diagnose.md
What diagnose does on Reprise
AI agents call diagnose to retrieve information from Reprise 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 |
|---|---|---|---|
original | string | — | The original image as an http(s) URL or a base64 data URL. |
reproduction | string | — | The reproduced image to compare, as an http(s) URL or a base64 data URL. |
Parameters from the server's own tool schema.
Why diagnose is rated Low
This tool performs comparison and analysis of image fidelity metrics (MAE, PSNR, exact-match percentage, stray pixels) between an original and reproduction. It retrieves and computes metrics without modifying, creating, deleting, or executing external operations. It is a read-only diagnostic function that returns data about the quality of a match.
From the tool's definition Tool name is 'diagnose' and description states it 'Score[s] how faithfully a reproduction matches an original' and 'Returns: A JSON string with the fidelity metrics' — purely a measurement/comparison operation with no side effects.
Attacks that exploit this kind of access
The rule that runs diagnose safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Reprise, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For diagnose, this is the rule to start with:
diagnose 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 Reprise, apply this rule, and every diagnose call is checked against it from then on.
Questions about diagnose
Score how faithfully a reproduction matches an original (MAE / PSNR / exact-match % / stray px). Returns: A JSON string with the fidelity metrics. It is categorised as a Read tool in the Reprise MCP Server, which means it retrieves data without modifying state.
diagnose accepts 2 parameters: original, reproduction. The full parameter table on this page comes from the server's own tool schema.
Register the Reprise MCP server in PolicyLayer and add a rule for diagnose: 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 Reprise. Nothing to install.
diagnose 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 diagnose 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 diagnose. 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.
diagnose is provided by the Reprise MCP server (https://tepesama-reprise-mcp.hf.space/gradio_api/mcp/http). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Reprise, and thousands of servers like it.
This server
Across the catalogue