ChuanTianML/prompt-for-me ↗★ 2
dsh-prompt-for-me
Context-aware next-message suggestions for the DeepSeek Harness composer, always reviewed by the user before sending.
安装
$
npx -p @deepseek-ai/dsh dsh plugin --profile web add github:ChuanTianML/prompt-for-me说明文档
阅读完整 README ↗Configuration
All generation limits are configurable in cordis.patch.yml:
| Field | Default | Meaning |
|---|---|---|
candidateCount | 3 | Suggestions required per batch. |
maxCandidateBytes | 4096 | UTF-8 limit per suggestion. |
maxDraftBytes | 32768 | UTF-8 limit for a draft or edited outcome. |
maxCurrentContextBytes | 65536 | JSON budget for current-session text. |
maxHistorySessions | 20 | Earlier sessions inspected. |
maxHistoryMessages | 100 | Earlier direct user prompts retained. |
maxHistoryBytes | 65536 | JSON budget for earlier prompts. |
maxLocalOutcomes | 50 | Browser-local suggestion outcomes retained. |
maxOutputTokens | 2048 | Auxiliary model output budget. |
timeoutMs | 30000 | Auxiliary model-call timeout. |
shortcut | Mod+Shift+Space | Portable Trigger, or disabled. |