feat(examples/appworld): add AppWorld RLM benchmark#32

Merged

glesperance merged 29 commits into

mainfrom

appworld-rlm

Jun 4, 2026

Commits on Jun 1, 2026

chore(examples/appworld): initial commit
glesperance
committed
feat(rlm-gepa): expose async evaluation adapter
glesperance
committed
docs(rlm): refresh generated skill guidance
glesperance
committed
feat(predict-rlm): support defaulted single-output SUBMIT
glesperance
committed
fix(examples/appworld): preserve local AppWorld venv interpreter path
glesperance
committed
refactor(examples/appworld): use host-bound AppWorld task completion
glesperance
committed
feat(examples/appworld): default AppWorld eval concurrency to 30
glesperance
committed
chore(predict-rlm): ignore nested Hermes state
glesperance
committed
docs(examples/appworld): refresh benchmark README formatting
glesperance
committed
chore(predict-rlm): rename agent guidance
glesperance
committed
chore: clarify tmux dashboard pattern for local runs
glesperance
committed
refactor(examples/appworld): lazily materialize default skill instructions
glesperance
committed
feat(examples/appworld): render ICL demos from official task data
glesperance
committed
chore(appworld): lower default concurrency
glesperance
committed
fix(examples/appworld): add eval progress diagnostics
glesperance
committed
fix(examples/appworld): avoid stderr pipe EOF hang
glesperance
committed
fix(predict-rlm): avoid async sandbox EOF stderr hang
glesperance
committed
fix(examples/appworld): expose dict kwargs to RLM tools
glesperance
committed
feat(rlm-gepa): support grouped AppWorld optimization
glesperance
committed
Fix AppWorld CodexLM launch routing
glesperance
committed
chore(predict-rlm): preserve supporting branch updates
glesperance
committed
docs(examples/appworld): update AppWorld blog post
glesperance
committed
docs(examples/appworld): document protected runs bundle
glesperance
committed
docs(examples/appworld): link hosted runs bundle
glesperance
committed
docs(examples/appworld): add runs bundle make targets
glesperance
committed
docs(examples/appworld): exclude macOS metadata from runs bundle
glesperance
committed

Commits on Jun 4, 2026

fix(examples/appworld): resolve AppWorld venv relative to example
glesperance
committed
feat(examples/appworld): add AppWorld RLM example
glesperance
committed
refactor(examples/appworld): remove defensive cleanup slop
glesperance
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat(examples/appworld): add AppWorld RLM benchmark#32

feat(examples/appworld): add AppWorld RLM benchmark#32
glesperance merged 29 commits into
mainfrom
appworld-rlm

Commits on Jun 1, 2026

chore(examples/appworld): initial commit

feat(rlm-gepa): expose async evaluation adapter

docs(rlm): refresh generated skill guidance

feat(predict-rlm): support defaulted single-output SUBMIT

fix(examples/appworld): preserve local AppWorld venv interpreter path

refactor(examples/appworld): use host-bound AppWorld task completion

feat(examples/appworld): default AppWorld eval concurrency to 30

chore(predict-rlm): ignore nested Hermes state

docs(examples/appworld): refresh benchmark README formatting

chore(predict-rlm): rename agent guidance

chore: clarify tmux dashboard pattern for local runs

refactor(examples/appworld): lazily materialize default skill instructions

feat(examples/appworld): render ICL demos from official task data

chore(appworld): lower default concurrency

fix(examples/appworld): add eval progress diagnostics

fix(examples/appworld): avoid stderr pipe EOF hang

fix(predict-rlm): avoid async sandbox EOF stderr hang

fix(examples/appworld): expose dict kwargs to RLM tools

feat(rlm-gepa): support grouped AppWorld optimization

Fix AppWorld CodexLM launch routing

chore(predict-rlm): preserve supporting branch updates

docs(examples/appworld): update AppWorld blog post

docs(examples/appworld): document protected runs bundle

docs(examples/appworld): link hosted runs bundle

docs(examples/appworld): add runs bundle make targets

docs(examples/appworld): exclude macOS metadata from runs bundle

Commits on Jun 4, 2026

fix(examples/appworld): resolve AppWorld venv relative to example

feat(examples/appworld): add AppWorld RLM example

refactor(examples/appworld): remove defensive cleanup slop