2 Commits

Author	SHA1	Message	Date
ZhenYi	94825316dc	fix(agent): extract JSON from model output even with leading text prefix ... ReAct loop was terminating early when the model returned: [Agent ran through N steps...] {"thought": "...", "action": {...}} The extract_json function only checked the string start or code fences. Now scans for { or [ at non-word positions and uses depth-counting to strip trailing text, allowing JSON buried anywhere in the response.	2026-04-24 13:17:06 +08:00
ZhenYi	93cfff9738	init	2026-04-15 09:08:09 +08:00