feat: add long-horizon agent eval project - #42

Open

zredlined wants to merge 11 commits into

mainfrom

codex/long-horizon-agent-evals

Commits on Aug 20, 2026

feat: add long-horizon agent eval project
zredlined
committed
5721cd0
docs: explain long-horizon eval architecture
zredlined
committed
5cd5456
docs: simplify eval setup diagram
zredlined
committed
8769f8a

Commits on Aug 21, 2026

fix: harden long-horizon eval tracing
zredlined
committed
f542b03
fix: bound long-horizon reviewer context
zredlined
committed
76f95f2

Commits on Aug 22, 2026

fix: rotate stalled challenger threads
zredlined
committed
3a35580
fix: abort on failed policy reloads
zredlined
committed
e5dfba4

Commits on Aug 24, 2026

feat: rotate challenger threads on a no-progress lull

zredlined
and
claude
committed
610db34
fix: preserve actionable challenger handoffs
zredlined
committed
64c3532

Commits on Aug 25, 2026

fix: make long-horizon results auditable
zredlined
committed
81b17e4
fix: abort scale on proposal preflight failure
zredlined
committed
27ebae1