AI agents use add_trial to create or update resources in Optuna MCP Server — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Optuna MCP Server environment.
Adding a trial creates new data in the study but does not irreversibly delete data or execute arbitrary code. It is a write operation that modifies the study's state. Severity is medium because misuse could pollute optimization results and waste computational resources, but the effect is reversible through study reset or deletion.
From the tool's definition Tool name 'add_trial' and description 'Add a trial to the study' indicate creation of new data (a trial record) within an optimization study. This is a reversible modification operation.
Documented attack patterns abuse exactly the kind of access add_trial gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Optuna MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for add_trial:
{
"version": "1",
"default": "deny",
"tools": {
"add_trial": {
"limits": [
{
"counter": "add_trial_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}add_trial stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Add a trial to the study. It is categorised as a Write tool in the Optuna MCP Server MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Optuna MCP Server MCP server in PolicyLayer and add a rule for add_trial: 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 Optuna MCP Server. Nothing to install.
add_trial is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the add_trial 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 add_trial. 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.
add_trial is provided by the Optuna MCP Server MCP server (optuna/optuna-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 26 Optuna MCP Server tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
26 Optuna MCP Server tools catalogued and risk-classified — across an index of 42,500+ MCP servers.