deixic
Log inRequest a demo

← Blog

Topic

Engineering

Use Deixic Code in the terminal, scripts, and CI with company context, and replay real sessions when changing a model, prompt, or method.

34 articles

SeriesMake your agents finish the job →Choosing and rolling out agents →Why did the agent do that? →Carry work between connected systems →

Lead article

· 8 min read

How we built an integration factory

How Deixic uses typed provider contracts, a shared Rust runtime, and executable verification to expand integrations without rebuilding the plumbing for each API.

Archive

All Engineering articles

Browse every article →
  1. When a passing test runs the wrong code4 min
  2. Stop letting GitHub agents rediscover bugs you already fixed4 min
  3. How we turned our coding agent into our own bugbot8 min
  4. When a source is unavailable, should the agent use a fallback?4 min
  5. Stopping an agent does not automatically undo its last tool call4 min
  6. The write succeeded. Why does the agent still see the old value?4 min
  7. How do I stop Stripe retries from doing the work twice?4 min
  8. A background job needs a result contract the agent can finish4 min
  9. Why does my AI agent keep calling the same tool?4 min
  10. How to debug an agent without copying every secret into its traces4 min
  11. How to replace a connector without changing what the agent means4 min
  12. An agent framework upgrade needs more than a passing import4 min
  13. When should the parent agent stop waiting for a missing child?4 min
  14. Two agents finished their parts. Why did one result disappear?4 min
  15. Your agent joined two correct facts from different moments4 min
  16. Why does an agent still quote a document after it was deleted?4 min
  17. Your agent found a green status. Did the work actually finish?4 min
  18. AI agent troubleshooting: find the question behind the failure4 min
  19. An expired sync token should trigger recovery, not an empty knowledge base4 min
  20. Your agent is under its limit. Why is the connector still throttled?4 min
  21. Why we built our own agent sandbox5 min
  22. Why one failed agent call becomes many provider requests4 min
  23. A tool timed out. Did the change still happen?4 min
  24. What happens when a tool schema changes halfway through an agent task?4 min
  25. A 1-second submit hid a 30-minute outage6 min
  26. The tool exists, but your agent cannot discover it4 min
  27. After the outage: making agent completion provable5 min
  28. Why does an agent choose the wrong tool when both names sound right?4 min
  29. When a tool response is too large, keep the fields the task needs4 min
  30. Your agent listed the first page. Did it finish the inventory?4 min
  31. Your AI coding agent is probably lying about its tests. It doesn’t know it is.3 min
  32. Why does the right document disappear when an agent adds a filter?4 min
  33. Why can my agent find a Notion page but not read its contents?4 min