create_prediction
Create a new prediction using either a model version (for community models) or model name (for official models)
This record as markdown: /tools/deepfates-mcp-replicate/create-prediction.md
What create_prediction does on Replicate MCP Server
AI agents invoke create_prediction to trigger actions in Replicate MCP Server. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.
Why create_prediction is rated High
Creating a prediction executes an AI model on external compute infrastructure. This is an Execute action because it triggers an external operation (model inference) whose effects and resource consumption depend on the arguments provided (model version/name and inputs). Misuse could result in high API costs or unintended model runs at scale.
From the tool's definition "Create a new prediction" - triggers external AI model inference operations on Replicate's infrastructure
Attacks that exploit this kind of access
The rule that runs create_prediction safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Replicate MCP Server, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For create_prediction, this is the rule to start with:
create_prediction stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Replicate MCP Server, apply this rule, and every create_prediction call is checked against it from then on.
Questions about create_prediction
Create a new prediction using either a model version (for community models) or model name (for official models). It is categorised as a Execute tool in the Replicate MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Replicate MCP Server MCP server in PolicyLayer and add a rule for create_prediction: 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 Replicate MCP Server. Nothing to install.
create_prediction is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the create_prediction 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 create_prediction. 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.
create_prediction is provided by the Replicate MCP Server MCP server (deepfates/mcp-replicate). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Replicate MCP Server, and thousands of servers like it.
Across the catalogue