AI Engineer · October 4, 2026

Durable human approval: resume agent workflows after a worker restart

video Agent Security AI Engineering

Durable human approval: resume agent workflows after a worker restart video thumbnail

Why it matters

Melanie Warrick models a human decision as an asynchronous signal to a durable workflow. External model and tool calls belong in activities, while replayable orchestration records progress and waits for approval without keeping a worker occupied. In the demonstration, a worker stops while an approval is pending; the approval is recorded and processing resumes after restart, while other orders can continue. This tests a worker interruption, not failure of every infrastructure component or the correctness of every possible approval integration.

My takeaway: Test a restart while approval is pending and while the approver responds. Correlate the signal with the exact order and requested action, define an expiry path, and make external writes idempotent so recovery cannot repeat a payment or publication.

Keep exploring

More curated notes connected through Agent Security and AI Engineering.

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.

Microsoft Security Blog · guide

AI vulnerability research: measure reproducible findings and completed fixes

Microsoft’s FORGE account describes the work between a model’s vulnerability claim and a useful repair: reusable builds, duplicate removal, reachability checks, project-specific verification, reproducible triggers and regression tests. Structured rejection reasons help improve later searches. The useful operational measure is the flow of findings that survive verification and reach a fix, rather than the number of candidates generated. Reported successful-case costs exclude parts of screening, failed attempts and human work, so they are not the total cost of operating this pipeline.

OpenAI News · framework

Frontier training safety cases: connect evidence to enforced pause and rollback controls

OpenAI proposes training-run safety cases combining alignment evaluations, containment and monitoring with explicit operational ownership. Concrete measures include immutable transcripts, held-out incident tests, checks for evaluation gaming, response deadlines and fail-closed monitoring. Independent internal challenge, leadership vetoes and tracking downstream uses support stopping a run and reversing affected work. The article describes recommendations still being implemented, rather than audited proof that every safeguard already operates or that residual risk has been eliminated.