AI Engineer · October 18, 2026

Agent workflow contracts: separate requested work from evidence of completion

video AI Engineering Agent Security Model Evaluation

Agent workflow contracts: separate requested work from evidence of completion video thumbnail

Why it matters

OpenProse expresses workflows as composable Markdown contracts. Its public examples distinguish input data from instructions, bound permitted work, and require evidence for factual claims. The presenters acknowledge that an agent can declare its own contract fulfilled. The experimental library distinguishes a completed evaluation from a successful task; independent review still has to inspect the actual artifacts.

My takeaway: Pin contract versions and state inputs, allowed effects, acceptance criteria, and required reviewers. Test a workflow with missing evidence, conflicting requirements, and instructions embedded in input data. Have an independent checker inspect outputs against the contract, record unmet conditions, and keep tool permissions enforced by the execution environment.

Keep exploring

More curated notes connected through AI Engineering and Agent Security.

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.