All pull requests

Pull requests

• refactor: return Results from tar pipe helpers in cache-store

  #175·sunwrobert opened Oct 31, 2026Contributor·3 of 5 tasks·

  ·
  7

  

• refactor: return Result from parquet merge and read instead of throwing

  #174·sunwrobert opened Oct 31, 2026Contributor·5 tasks done·

  ·
  1

  

• refactor(cli): return effect/Result from argument and config validation

  #173·sunwrobert opened Oct 31, 2026Contributor·3 of 5 tasks·

  ·
  1

  

• refactor: return a Result from mirror-vgi-bench-media readOptions

  #172·sunwrobert opened Oct 31, 2026Contributor·5 tasks done·

  ·
  1

  

• fix: sandbox agent pins and origin targets, agent payload plugin merge, Access-scoped redirect refusal

  #171·LukasParke opened Oct 30, 2026Contributor·3 of 5 tasks·

  ·
  4

  

• deps: bump the dependencies group across 1 directory with 16 updates

  #162·dependabot[bot] opened Oct 29, 2026Bot·

  ·
  1

• feat(agent-cli): report tool-call progress events for CLI agents and the swe-atlas loop

  #148·jamespsterling opened Oct 13, 2026Contributor·3 of 5 tasks·

  ·
  20

  

• refactor(benchmarks): retire Tau3 Banking from Kepler (2/2)

  #140·ayush-or opened Oct 9, 2026Contributor·

  ·

• feat(results): offload inline base64 payloads to an artifact sink

  #138·abhinav-pola opened Oct 8, 2026Contributor·5 tasks done·

  ·
  15

  

• refactor(benchmarks): retire Harbor-backed Kepler implementations (1/2)

  #113·abhinav-pola opened Sep 23, 2026Contributor·4 of 5 tasks·

  ·
  4

  

• feat(recovery-bench): add Recovery-Bench failed-state recovery benchmark

  #110·louisgv opened Sep 23, 2026Member·5 tasks done·

  ·
  22

  

• fix(sandbox): match Modal region pins through the openrouter/ routing prefix

  #99·devin-ai-integration[bot] opened Sep 18, 2026Bot·3 of 5 tasks·

  ·
  1

  

• feat(probably-decisions): add Probably-judged Sentinel decision benchmark

  #96·talos opened Sep 18, 2026·5 tasks done·

  ·
  1

  

• fix(harness): treat HTTP 408 timeouts as retryable and score exhausted timeouts as Skipped

  #83·devin-ai-integration[bot] opened Sep 11, 2026Bot·3 of 5 tasks·

  ·
  1

• fix(cli): forward solver config to all model benchmarks and improve error resilience

  #77·devin-ai-integration[bot] opened Sep 8, 2026Bot·3 of 5 tasks·

  ·
  1

• feat(agent-cli): add ori code harness for terminal-bench

  #76·devin-ai-integration[bot] opened Sep 8, 2026Bot·3 of 5 tasks·

  ·
  1

  

• feat: persist completed sample outcomes so interrupted runs resume instead of restarting

  #48·jamespsterling opened Aug 21, 2026Contributor·4 of 5 tasks·

  ·
  33

  

• fix(search): resolve server-tool generation ids

  #32·ayush-or opened Aug 14, 2026Contributor·

  ·

• feat(benchmarks): chess — full games vs Stockfish, first-party original

  #4·LukasParke opened Aug 5, 2026Contributor·

  ·
  2

• feat(benchmarks): EvalManifestV1 — unified eval description schema + first-party manifests

  #3·LukasParke opened Aug 5, 2026Contributor·

  ·
  2

• feat(benchmarks): custom_eval — declarative user evals as run-config data

  #2·LukasParke opened Aug 5, 2026Contributor·

  ·
  2

• feat(harness): durable per-(sample, epoch) result store + fiber-scoped request context

  #1·LukasParke opened Aug 5, 2026Contributor·

  ·
  3