Teknium 7a70cbd0db fix(evals): reasoning-replay treats Anthropic refusals like xAI 403s ...

Anthropic answers a probe that asks about "earlier reasoning" with a 200
content_filter refusal and no usage block, which left cross-turn cells
without a baseline. Both refusal shapes now rerun the cell with neutral
wording, and opaque cells pick the newest turn-1 capture per position so a
partial rerun does not mix captures.

2026-10-10 20:49:30 -07:00

..

botmode-dm-delivery

…

botmode-dm-matrix

…

browser_use

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

codebase_navigability

ci: remove unused variables (linter)

2026-10-10 15:20:14 +02:00

codex_echo

…

compaction

ci: remove unused variables (linter)

2026-10-10 15:20:14 +02:00

core_tool_deferral

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

dashboard_auth

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

delegation_group_schema

…

desktop_bug_campaign

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

desktop_mcp_oauth

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

desktop-pane-fixture

…

gateway

…

gateway_completion

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

gateway_failure_ownership

ci: remove unused variables (linter)

2026-10-10 15:20:14 +02:00

gateway_status_render

…

liveness

…

native_compaction

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

openrouter_pkce_ab

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

plugin_isolation

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

postmortem

ci: remove unused variables (linter)

2026-10-10 15:20:14 +02:00

prompt_footprint

…

provider_fallback

…

provider_wire

ci: fix PLC0206, PLC0414 linter errors

2026-10-10 15:20:14 +02:00

providers

…

readtool

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

reasoning_replay

fix(evals): reasoning-replay treats Anthropic refusals like xAI 403s

2026-10-10 20:49:30 -07:00

session_search_schema

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

subagent_process_handoff

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

token_accounting

ci: remove unused variables (linter)

2026-10-10 15:20:14 +02:00

tool_search

ci: remove unused variables (linter)

2026-10-10 15:20:14 +02:00

toolperf_abeval

ci: fix PLC0206, PLC0414 linter errors

2026-10-10 15:20:14 +02:00

update_pipeline

…

update_streaming

…

webhook_auth

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

__init__.py

…

acp_empty_session_wire.py

…

anthropic_proxy_thinking_replay.py

…

api_delegation_http_probe.py

…

api_delegation_sync_probe.py

…

approval_deny_dispatch.py

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

auth_pool_controls.py

…

auxiliary_resource_exhausted.py

…

cli_deferred_notice.py

…

cli_fallback_add_picker_error.py

…

codex_masked_replay_review.py

…

completion_backlog_probe.py

…

cron_error_diagnostics.py

…

cron_timeout_fork_race.py

…

delivery_flood_wire.py

…

desktop-reap-clock-pressure.cjs

…

fanout_resource_bench.py

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

gemini_type_array_probe.py

…

goal_command_parity.py

…

heartbeat_idle_wire.py

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00

kanban_graph_identity.py

…

kanban_pr_acceptance_live.py

…

kanban_scope_probe.py

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

key_cmd_picker_live.py

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

mcp_device_flow.py

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

output_caps_local_capture.py

…

output_caps_scope.md

…

output_caps_surfaces.py

…

process_result_receipt_probe.py

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

session_snapshot_removal.py

…

slack_stream_wire_contract.py

ci: fix remaining ASYNCXXX linter errors

2026-10-10 15:20:14 +02:00

update_check_ssh_pty.py

…

update_obligation_identity.py

…

update_unit_client_budget.py

ci: fix PLW1509, PLW1510, PLW2101 linter errors

2026-10-10 15:20:13 +02:00

vault_fill_live_e2e.py

ci: apply all automatic linter fixes available

2026-10-08 17:22:54 +02:00