Self-learning habit engine for DeepSeek Harness
Max-Null/dsh-habit · skills
Self-learning habit engine for the DeepSeek Harness — observes user-correction signals from session events, judges habits with a low-cost model on threshold, and settles candidates behind a two-level human gate.
★ 2GitHub stars
0Forks
2026-08-29Last updated
TypeScriptLanguage
—License
Key features
- snapshot() → candidates (newest first)
- Deterministic observation, LLM on demand: correction detection is a fixed phrase list + length cap (task descriptions are not corrections)
- confirm(id) / discard(id) → first-level human gate
- (the second gate is dsh-memory's own suggested→auto confirmation)
- the LLM only runs when a session accumulates enough signals.
Install command
dsh plugin --profile web add github:Max-Null/dsh-habit