Deepseek API
Hi everyone,
I'm running into a persistent issue with my OpenClaw (formerly Clawdbot) setup on Discord. The bot responds with "Message ordering conflict - please try again. If this persists, use /new to start a fresh session." every time I send a message, even after using /new, /reset, restarting the bot, and clearing sessions.
Setup Details:
Version: 2026.1.29 (from config meta)
OS: Windows 10/11 (running in PowerShell)
Primary Model: deepseek/deepseek-chat (custom provider via models.providers with OpenAI-compatible API)
Channel: Discord (enabled in config, with a bot token)
Config is mostly default from onboarding, with custom DeepSeek integration and some fallbacks (e.g., google-antigravity, opencode models).
Workspace: C:\Users[myuser]\clawd
What I've Tried:
Ran openclaw restart and openclaw stop + manual process kill + openclaw start.
Deleted all files in ~/.openclaw/sessions/ (or Windows equivalent).
Removed potentially buggy models like "opencode/glm-4.7-free" from allowlist and fallbacks.
Ran openclaw doctor - no errors reported.
Verified DeepSeek API connection works via manual test script (responses come back fine).
Re-onboarded with openclaw onboard.
Logs show some "role" validation errors tied to opencode models, but even after removing them, the issue persists. Has anyone seen this? Could it be a session corruption bug from GLM or something Discord-specific?
I'm running into a persistent issue with my OpenClaw (formerly Clawdbot) setup on Discord. The bot responds with "Message ordering conflict - please try again. If this persists, use /new to start a fresh session." every time I send a message, even after using /new, /reset, restarting the bot, and clearing sessions.
Setup Details:
Version: 2026.1.29 (from config meta)
OS: Windows 10/11 (running in PowerShell)
Primary Model: deepseek/deepseek-chat (custom provider via models.providers with OpenAI-compatible API)
Channel: Discord (enabled in config, with a bot token)
Config is mostly default from onboarding, with custom DeepSeek integration and some fallbacks (e.g., google-antigravity, opencode models).
Workspace: C:\Users[myuser]\clawd
What I've Tried:
Ran openclaw restart and openclaw stop + manual process kill + openclaw start.
Deleted all files in ~/.openclaw/sessions/ (or Windows equivalent).
Removed potentially buggy models like "opencode/glm-4.7-free" from allowlist and fallbacks.
Ran openclaw doctor - no errors reported.
Verified DeepSeek API connection works via manual test script (responses come back fine).
Re-onboarded with openclaw onboard.
Logs show some "role" validation errors tied to opencode models, but even after removing them, the issue persists. Has anyone seen this? Could it be a session corruption bug from GLM or something Discord-specific?
The AI that actually does things. Emails, calendar, home automation - all from your favorite chat app. New shell, ready to help. The lobster way.
83,815Members
Resources
Similar Threads
Was this page helpful?
