R094 · Agent principles and patterns › Long-running agents and durable execution

CodeClimb: an overnight self-improving coding agent on durable workflows

Read the original

durable-executionself-improvementcost-controlevaluationcoding-agents

Source: Paul Iusztin's first-hand build notes on LinkedIn (14 to 15 November 2026) about CodeClimb, his agent that writes, evaluates and improves its own code overnight, built on Inngest. Not independently verified.

The setup

CodeClimb runs a coding agent (Pi) in rounds: write a solution, score it on a dev split, keep it only if it beats the previous best by more than a noise margin, repeat. The first target was a GraphRAG pipeline over a custom ontology with explicit metric goals (extraction F1 0.90, relation F1 0.40, graph-merge precision and recall 0.98, single-hop recall 0.70, multi-hop completeness 0.30).

Architecture

Results so far

The first overnight GraphRAG run cost about $10 across five experiments. A later post reports about 75% on graph extraction on autopilot, but the targets were not yet all met.

Lessons the author reports

Why it matters

It is a compact reference design for unattended agent work: durable execution, hidden test data, external budget enforcement, scoped credentials and a noise margin for acceptance. Each piece maps to a known failure of overnight agents.

Review checklist

  1. Can our unattended agents see the data they are graded on?

  2. Are budgets enforced by something the agent cannot edit?

  3. Do crashes resume from checkpoints, or restart?

  4. Do we keep a record of rejected attempts so the agent does not repeat them?

  5. Is improvement judged against a noise margin?

R084 (always-on agent primitives), R035 (durable execution), R086 (long-horizon agent benchmarks), R039 (noise floors), R023.

Connections

See R094 on the knowledge map

Part of the Research notes on nowucca.com. Figures and claims are the source's own unless the summary says otherwise.