vultr/openclaw
1
0
Fork 0
mirror of https://github.com/openclaw/openclaw.git synced 2026-04-17 12:11:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(discord): raise thread title max tokens for reasoning models

Browse Source 
When the simple-completion model selected for thread-title generation is a
reasoning model (e.g. MiniMax M2, Claude thinking models, OpenAI o-series),
the 24-token output budget is entirely consumed by the internal thinking
block before any user-visible text is emitted. extractAssistantText then
returns an empty string, generateThreadTitle returns null, and the
auto-thread rename is silently skipped while the feature appears to do
nothing.

Raise DISCORD_THREAD_TITLE_MAX_TOKENS to 512 so there is enough headroom
for a short thinking pass plus the 3-6 word title output. The generous
ceiling only matters when the provider actually reasons; non-reasoning
models still emit a short title and stop early at end-of-sequence.

Verified live against a MiniMax M2 reasoning model served through an
Anthropic-compatible API endpoint: before the fix, the rename never fired;
after the fix, the thread is renamed with a concise generated title.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Hana Chang
2026-04-10 15:32:10 +08:00
committed by Peter Steinberger
parent 13821fd54b
commit 537479f5b0
2 changed files with 7 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
extensions/discord/src/monitor/thread-title.generate.test.ts
View File

@@ -172,7 +172,7 @@ describe("generateThreadTitle", () => {
).toContain("Channel description: Deploy updates and incident notes");
expect(completeWithPreparedSimpleCompletionModelMock.mock.calls[0]?.[0]?.options).toEqual(
expect.objectContaining({
maxTokens: 24,
maxTokens: 512,
}),
);
expect(

7
extensions/discord/src/monitor/thread-title.ts
View File

@@ -10,7 +10,12 @@ const DEFAULT_THREAD_TITLE_TIMEOUT_MS = 10_000;
const MAX_THREAD_TITLE_SOURCE_CHARS = 600;
const MAX_THREAD_TITLE_CHANNEL_NAME_CHARS = 120;
const MAX_THREAD_TITLE_CHANNEL_DESCRIPTION_CHARS = 320;
const DISCORD_THREAD_TITLE_MAX_TOKENS = 24;
// Budget generous enough to cover reasoning-model thinking tokens plus the
// short text output. Lower values (e.g. 24) starve reasoning models of output
// capacity: the entire budget is consumed by the thinking block before any
// text is emitted, so extractAssistantText returns empty and the rename is
// silently skipped.
const DISCORD_THREAD_TITLE_MAX_TOKENS = 512;
const DISCORD_THREAD_TITLE_SYSTEM_PROMPT =
"Generate a concise Discord thread title (3-6 words). Return only the title. Use channel context when provided and avoid redundant channel-name words unless needed for clarity.";