Retry a failed or timed-out Serverless job. Only works for jobs with FAILED or TIMED_OUT status. The previous output is removed and the job is requeued.
AI agents invoke retry-job to trigger actions in RunPod 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.
This tool re-executes a serverless job, which constitutes triggering an external operation. While it also removes previous output (a minor destructive side effect), the primary action is re-running/re-queuing a job execution. The blast radius is medium since it could trigger unintended compute workloads or repeated operations, but it is scoped to already-failed jobs.
From the tool's definition 'Retry a failed or timed-out Serverless job' and 'the job is requeued' — triggers re-execution of an external operation
Documented attack patterns abuse exactly the kind of access retry-job gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and RunPod MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for retry-job:
{
"version": "1",
"default": "deny",
"tools": {
"retry-job": {
"limits": [
{
"counter": "retry-job_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}retry-job 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.
Free to start. No card required.
Retry a failed or timed-out Serverless job. Only works for jobs with FAILED or TIMED_OUT status. The previous output is removed and the job is requeued. It is categorised as a Execute tool in the RunPod MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the RunPod MCP Server MCP server in PolicyLayer and add a rule for retry-job: 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 RunPod MCP Server. Nothing to install.
retry-job 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 retry-job 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 retry-job. 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.
retry-job is provided by the RunPod MCP Server MCP server (runpod/runpod-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 36 RunPod MCP Server tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
36 RunPod MCP Server tools catalogued and risk-classified — across an index of 42,500+ MCP servers.