memory-to-skill

1<!--
2  memsearch-dsh memory-to-skill skill body.
3  Registered at plugin load by plugins/dsh/index.js via ctx.skills.register()
4  with metadata (name: memory-to-skill, description, whenToUse) supplied in code.
5-->
6 
7 
8You manage MemSearch's **procedural memory**: skills distilled from the work you
9repeat — a third layer beside the daily journals (episodic) and PROJECT.md /
10USER.md (semantic). State once that this is MemSearch skill distillation, not
11the host agent's built-in skills system.
12 
13Stages: **0** memory journals → **1** candidate (`.memsearch/skill-candidates/`,
14a git-tracked store that keeps evolving) → **2** installed (an agent skill dir).
15Candidates are never installed automatically; installing is always a human step.
16User requests may stop at candidate creation/review, or continue to installation
17in the same turn after explicit approval; match the requested stage.
18 
19The `plugins.<platform>.memory_to_skill.*` config key prefix and the install-path
20notes are platform-specific — see your platform reference file:
21 
22- Claude Code → `references/claude-code.md`
23- Codex → `references/codex.md`
24- OpenClaw → `references/openclaw.md`
25- OpenCode → `references/opencode.md`
26- DeepSeek Harness → `references/dsh.md`
27 
28## Intent routing
29 
30- "make/turn this into a skill", "from what we just did" → **A. Capture now**.
31- "what skills / review candidates / install X" → **B. Review & install**.
32- "mine my history / find recurring workflows" → **C. Distill from history**.
33…