Tag
A discussion of why per-run budgets and retry limits in AI agent systems fail to catch slow, task-level repetition, and a proposal to scope limits to tasks with persistent identifiers, decay-window counters, and error paths that share the remaining budget.
Orthogonal launched a service that helps always-on AI agents avoid wasting paid tool calls by checking usage after ambiguous timeouts or truncated responses before retrying, letting agents autonomously find, use, and pay for data services from many companies. The launch was highlighted alongside OpenAI's Dots integration for autonomous lead-generation pipelines.
The author shares lessons from putting an AI agent with write access into production, explaining that retry logic inside the agent loop fails when the process dies. They advocate for treating side-effectful tool calls as durable background jobs with idempotency keys.
A developer shares a lesson learned about capping tool retries per session rather than per call to prevent runaway costs from flaky endpoints, recommending a per-session retry budget that fails loud when exhausted.