Articles
AI agents
- AI agents (13)
- AI Coding Agents (1)
- AI delivery (88)
- AI Delivery Programme (12)
- AI Workflow Audit (2)
- Architecture Review (2)
- Laravel (1)
- MVP to Production (1)
-
How do I notice when an AI agent or automation reports success but got it wrong?
A success status only says the steps finished. Catch quiet failures with outcome checks in code, expected-volume alarms and a heartbeat per trigger, and…
-
How should an AI agent hold and use the credentials for the business systems it acts on?
Give each agent its own identity and a named sponsor, let the tool layer fetch a narrow, short-lived credential per action, keep secrets out…
-
How do I make an AI agent stop and hand the conversation to a person instead of guessing?
Decide the stop points before launch, check them in code on every message, record who owns each conversation so the agent stays silent during…
-
How do I know my agent’s answers are right in production without an answer key?
Without labelled answers, build your own answer key in layers: code checks against systems of record, a graded random sample, and a model grader…
-
How do I test that my AI agent calls the right tool with the right arguments without touching real systems?
Put the code that executes tools behind a seam, swap in recording fakes, assert the tool name and the exact arguments, repeat each scenario,…
-
Does this task need an AI agent, a simple automation or ordinary software?
Decide step by step: rules go to ordinary code or a simple automation, interpretation to one checked model step, and only a path that…