AI Engineer · October 26, 2026

Harbor evaluations: validate the verifier before measuring the agent

video Model Evaluation AI Engineering AI Red Teaming

Harbor evaluations: validate the verifier before measuring the agent video thumbnail

Why it matters

Xander Song’s Harbor workshop packages a task with instructions, a seeded environment and scripts that grade the resulting state. Oracle solutions check the setup before model calls, while fresh trials capture variation. A migration example checks schema changes; a later prompt experiment compares tool use. Passing an oracle alone does not prove the grader rejects wrong solutions.

My takeaway: Create a small task with explicit outcomes, a known-good solution and deliberately incorrect outputs. Verify that the grader accepts and rejects the right cases before testing agents. Use fresh environments, repeated trials and protected grading artifacts. Save trajectories to explain changes, and measure task correctness separately from tool-call count and cost.

Keep exploring

More curated notes connected through Model Evaluation and AI Engineering.

OpenAI News · framework

Pacing model development in an era of cyber-critical capabilities

OpenAI says preliminary evidence that Astra may meet its Critical cybersecurity threshold led it to pause frontier reinforcement-learning work for two weeks and keep its largest planned run on hold. New safeguards include stronger workload and network isolation, continuous boundary testing, token-level monitoring that escalates suspicious tool activity, and broader alignment checks for deception, reward hacking, and unauthorized access.

OWASP GenAI Security Project · guide

OWASP Top 10 for Agentic Applications for 2026

OWASP's community guide organizes agentic-system risk into ten categories, including goal hijacking, tool misuse, identity and privilege abuse, memory poisoning, insecure inter-agent communication, cascading failures, and rogue-agent behavior. It provides a shared taxonomy and mitigation starting point rather than a certification checklist or evidence that a deployed system is secure.

OpenAI News · framework

A blueprint for democratic governance of frontier AI

OpenAI proposes a three-part U.S. frontier-AI governance model: harmonize emerging state safety laws into a federal baseline, strengthen CAISI as an evaluation and standards institution, and coordinate a broader resilience program. Proposed controls include severe-risk evaluations, transparency reports, independent audits, safety-incident reporting, model-weight security, whistleblower protection, and periodic technical assessments.