Operated by LocalMark

LocalMark AgentOps Advisor

Inspect a supplied Agent Card, explain version-aware interface repairs and compare meaningful changes against an optional previous card. No live probe or ownership verification.

Limited free AI capacity

Try a task

Supply card and optionally previous_card, at most 32 KiB each. The local report owns the repair example and change list; AI explains the findings without executing a task.

Start with an example Task input (JSON)

Text and structured input only. No uploaded files, PDF parsing, external URLs to fetch or executable code.

I authorize this input to be sent to OpenAI for this task.

AI requests send the supplied input to OpenAI. LocalMark does not store request content or results. OpenAI may retain abuse-monitoring data under its API policy. Do not submit passwords, API keys, personal records or confidential material.

One AI request per minute and five per UTC day per client. Shared capacity limits also apply. A preview does not call OpenAI.

Result

Your result will appear here.

No task has been run.

Connect another agent

Read the Agent Card ↗ · Request formats and limits

A2A JSON-RPC: https://agentops.localmark.ai/a2a
REST: POST https://agentops.localmark.ai/api/run
MCP collection: https://agents.localmark.ai/mcp

Sending a task requires explicit allow_model_processing: true. Card retrieval, protocol task lookup and local previews do not spend model allowance.