Debug infinite tool loop in the bounded research agent

Do not start with a speculative fix for the bounded research agent. Force the failure on purpose, save the before output, then change one cause at a time. Retries are allowed only for transient conditions—not for bad input that will fail forever on agents-in-code.

for (let attempt = 0; attempt <= maxRetries; attempt++) {
  try { return await operation(signal); }
  catch (error) {
    const failure = classify(error);
    if (!failure.retryable || attempt === maxRetries) throw failure;
    await sleep(jitteredBackoff(attempt), signal);
  }
}
throw new Error("unreachable");

Expected evidence: The agent searches once, cites kb://returns, and stops before MAX_AGENT_STEPS.. If you cannot reproduce on demand, you do not yet control the failure mode for agents-in-code.

After repair, rerun the exact reproduction command. Keep the failing fixture as a regression seed for the observability page. For the bounded research agent over a fixture KB, remember the claim you are restoring: agent stops at max_steps; every claim cites a KB id or abstains.

Store a one-command reproduction for: infinite tool loop, or final answer with no citations after search. The command should use fixture KB with 4 notes + max_steps=3 or a minimal mutant of it. Paste the failing output into notes/failure-before.txt (or your shell scrollback as copied text). After the fix, paste notes/failure-after.txt and keep both.

Retries belong only on transient faults. If the failure is bad input, a bad allowlist, or a logic bug in the bounded research agent, retrying will amplify cost without repairing trust around search a local knowledge base then produce a cited answer within a step budget.

Classify the failure as prevent, detect, contain, or recover—using this lab’s language, not a generic poster. For agents-in-code, the first fix should usually be detect+prevent at the boundary, because infinite tool loop, or final answer with no citations after search is cheaper to stop early than to explain in production prose.

At the debugging stage for agents-in-code, the job is narrower than finishing a product demo. You are creating one progressive evidence piece about fixture KB with 4 notes + max_steps=3 that later pages inherit without redefining success. Keep that fixture small enough to inspect by hand, keep outputs copy-pasteable as text, and refuse to narrate this baseline as if it were a production SLA: single-shot LLM answer without search.

For this page specifically, success looks like before/after evidence for the characteristic failure while still centering the user decision to search a local knowledge base then produce a cited answer within a step budget. If you cannot point to a file, command, or assertion that proves that for the bounded research agent, stay on this page instead of advancing.

Glossary: tool · Glossary: structured output · Cheatsheet: production ops signals

Previous · Next