pj/messageboardbench

0 Fork 0

mirror of https://github.com/priyanshujain/messageboardbench.git synced 2026-10-02 11:07:07 +00:00

T

pj e94b75c6f5 archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

.agents/skills/experiment-analysis

archive old experiments and publish swe series

2026-09-25 12:34:44 +05:30

archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

src/messageboardbench

archive historical results and add token budget logs

2026-09-25 22:07:26 +05:30

.env.example

archive old experiments and publish swe series

2026-09-25 12:34:44 +05:30

.gitignore

archive old experiments and publish swe series

2026-09-25 12:34:44 +05:30

.python-version

Repo skeleton, pinned environment, free smoke test

2026-08-31 21:14:58 +05:30

AGENTS.md

archive old experiments and publish swe series

2026-09-25 12:34:44 +05:30

compose.team.yaml

Checkpoint experiments through SWE pilot v2

2026-09-15 15:46:10 +05:30

compose.yaml

Repo skeleton, pinned environment, free smoke test

2026-08-31 21:14:58 +05:30

EXPERIMENT.md

archive old experiments and publish swe series

2026-09-25 12:34:44 +05:30

justfile

Checkpoint experiments through SWE pilot v2

2026-09-15 15:46:10 +05:30

LICENSE

Repo skeleton, pinned environment, free smoke test

2026-08-31 21:14:58 +05:30

pyproject.toml

Checkpoint experiments through SWE pilot v2

2026-09-15 15:46:10 +05:30

README.md

archive old experiments and publish swe series

2026-09-25 12:34:44 +05:30

uv.lock

Checkpoint experiments through SWE pilot v2

2026-09-15 15:46:10 +05:30

README.md

messageboardbench

Research harness for studying whether coding agents publish, receive, and adopt evaluator-gaming methods across shared scratchpads and message boards.

The current study runs three OpenRouter models on the same 20 ImpossibleBench SWE tasks under isolated, scratchpad, board, feedback, internet, and token-awareness conditions.

See the current experiment, scenario series, experiment index, and setup guide.

Current raw runs live in logs/. Historical local experiments and logs stay in ignored archive/.