文件历史

提交图

5 次代码提交

作者 SHA1 备注 提交日期
Asim Aslam 205d6a55a3 Respect provider retry-after backoff hints (#3309)
Co-authored-by: Codex <codex@openai.com>
2026-06-28 23:57:17 +01:00
Asim Aslam 6da799d64f Classify AI retry failures (#3238)
Co-authored-by: Codex <codex@openai.com>
2026-06-28 10:42:37 +01:00
Asim Aslam 30f53e42a7 Expose model retry attempt metadata (#3207)
Co-authored-by: Codex <codex@openai.com>
2026-06-28 00:56:21 +01:00
Asim Aslam a8ea60120e agent: make model retries opt-in and harden retry backoff (#3021)
Follow-up to #3017. A Generate runs the whole tool-execution turn, so
auto-retrying it re-runs already-executed (possibly side-effecting) tool
calls. Default ModelMaxAttempts 3 -> 1: retries are now opt-in via
ModelRetry. The timeout stays as a safety net.

Also harden ai.GenerateWithRetry: always back off between retries
(exponential, capped at 30s, default 200ms if unset) so an opt-in retry
can't busy-loop the provider even with Backoff=0.

Verified: go build, go test -race ./agent/... ./ai/, golangci-lint.

Co-authored-by: Claude <noreply@anthropic.com>
2026-06-24 12:42:15 +01:00
Asim Aslam ab2c092693 Add agent model retry and timeout resilience (#3017) 2026-06-24 12:30:58 +01:00