haosenwang1018 8de2f8d588 fix: preserve reasoning_content in message sanitization for thinking models ...

_sanitize_messages strips all non-standard keys from messages, including
reasoning_content. Thinking-enabled models like Moonshot Kimi k2.5
require reasoning_content to be present in assistant tool call messages
when thinking mode is on, causing a BadRequestError (#1014).

Add reasoning_content to _ALLOWED_MSG_KEYS so it passes through
sanitization when present.

Fixes #1014

2026-02-24 04:21:55 +08:00

..

agent

improve agent reliability: behavioral constraints, full tool history, error hints

2026-02-23 09:13:08 +00:00

bus

feat(slack): isolate session context per thread

2026-02-23 10:23:55 +00:00

channels

Merge remote-tracking branch 'origin/main' into pr-1029

2026-02-23 14:01:43 +00:00

cli

fix(heartbeat): suppress progress messages and deliver agent response to user

2026-02-23 13:45:09 +00:00

config

improve agent reliability: behavioral constraints, full tool history, error hints

2026-02-23 09:13:08 +00:00

cron

Merge branch 'main' into pr-763

2026-02-20 08:04:48 +00:00

heartbeat

fix: remove extra blank line

2026-02-23 13:59:47 +00:00

providers

fix: preserve reasoning_content in message sanitization for thinking models

2026-02-24 04:21:55 +08:00

session

fix(session): get_history uses last_consolidated cursor, aligns to user turn

2026-02-23 16:57:08 +00:00

skills

fix(cron): fix timezone display bug, add tz validation and skill docs

2026-02-17 08:30:52 +00:00

templates

refactor: replace cron usage docs in TOOLS.md with reference to cron skill

2026-02-23 18:28:09 +00:00

utils

feat: redesign memory system — two-layer architecture with grep-based retrieval

2026-02-12 15:02:52 +00:00

__init__.py

release: v0.1.4

2026-02-18 14:31:26 +00:00

__main__.py

🐈nanobot: hello world!

2026-02-01 07:36:42 +00:00