Cognition runs Fable 5 inside Devin for eight-hour autonomous engineering
Cognition reported that Fable 5 could work for eight hours unattended inside Devin while making real engineering progress, versus earlier models drifting after minutes or roughly an hour. Some Fable-backed capabilities were already in product: Devin could monitor production or Slack, enter issues without being tagged, and independently triage incidents.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why it moved the index
Magnitude 57 reflects a documented jump to overnight autonomous software work plus proactive production monitoring in an operational agent product. Confidence 90 reflects a dated attributable account from the company deploying the model, with concrete duration, benchmark and product-behavior details.
Assessment history
-
R1
Toward 57 · confidence 90
New distinct operational autonomy evidence: Cognition documented overnight Fable 5 work and proactive deployed behavior inside Devin.
12 Aug 2026
Share this page
-
DoomBench assesses “Cognition runs Fable 5 inside Devin for eight-hour autonomous engineering” as evidence moving toward doom, with magnitude 57 and confidence 90 out of 100 in the autonomy and agency category.
-
The DoomBench assessment of “Cognition runs Fable 5 inside Devin for eight-hour autonomous engineering” is based on reporting from Claude by Anthropic and records the editorial rationale, source quality, attribution, and revision history.
-
DoomBench summarizes “Cognition runs Fable 5 inside Devin for eight-hour autonomous engineering” as follows: Cognition reported that Fable 5 could work for eight hours unattended inside Devin while making real engineering progress,...
https://www.doombench.com/news/cognition-runs-fable-5-inside-devin-for-eight-hour-autonomous-engineering-2026-07-10