7 Commits
Author SHA1 Message Date
Ahmed Allam 4a455b1e62 fix(core): recover from hallucinated tool names instead of ending the scan
A tool call for a name Strix does not register raised ModelBehaviorError
from the SDK turn resolver, which nothing retries: the root agent's raise
tore down the whole scan and a sub-agent died before its status was set.
Opt into the SDK's tool_not_found_behavior="return_error_to_model" so the
unknown call comes back as a tool result and the agent self-corrects.

The setting landed in openai-agents 0.19.0, which requires openai>=2.45,
so both pins move.
2026-08-04 06:14:54 +03:00
devin-ai-integration[bot]andGitHub d4e58b2cd0 fix(llm): pass LLM_EXTRA_HEADERS through ModelSettings so they reach the agent loop (#937) 2026-07-29 19:38:06 -07:00
27f9750cdc feat(llm): enable Bedrock/Anthropic prompt caching for Claude models (#772)
Co-authored-by: Sean Turner <sean.turner@zerohash.com>
Co-authored-by: Ahmed Allam <ahmed39652003@gmail.com>
2026-07-26 17:12:57 -07:00
Ahmed AllamandAhmed Allam 3bb95ab43d fix(llm): add per-turn model request timeout so stalled streams fail fast and retry 2026-07-17 18:40:23 -07:00
devin-ai-integration[bot]andGitHub 899e07d3a2 fix(core): bound per-agent image memory (proactive budget + inherited-context scrub) (#779) 2026-07-15 18:13:42 -07:00
alex sandGitHub 22d327d21f feat(settings): add force_required_tool_choice to LlmSettings (#730)
feat(inputs): implement logic for required tool choice based on model

test(inputs): add tests for force_required_tool_choice behavior

test(runner): update tests to include force_required_tool_choice in settings
2026-07-10 18:36:33 -04:00
Rome ThorstensonandGitHub 777005a42b fix: stop gracefully with resume hint on persistent RateLimitError (#261) (#593) 2026-06-29 07:31:54 -07:00