openclaw - 💡(How to fix) Fix [Question/Roadmap] memory-wiki — Is there a plan for populating concepts/entities/syntheses folders? [1 participants]

Context

I'm using OpenClaw v2026.4.11 (upgrading to v2026.4.15 shows the same behavior) with the memory-wiki plugin in unsafe-local vault mode. The plugin scaffolds the following folder layout at my vault root:

OpenClaw/
├── sources/       (559 files, actively populated by nightly import)
├── reports/       (6 files: claim-health, contradictions, lint, low-confidence, open-questions, stale-pages)
├── concepts/      (only index.md with "No concepts yet")
├── entities/      (only index.md with "No entities yet")
├── syntheses/     (only index.md with "No syntheses yet")
├── _attachments/  (empty)
└── _views/        (empty)

`openclaw wiki status` confirms:

``` Pages: 559 sources, 0 entities, 0 concepts, 0 syntheses, 6 reports ```

`openclaw wiki doctor` reports `healthy`.

What I've checked

• `openclaw wiki --help` — no sub-command for `concepts` / `entities` / `syntheses` extraction or generation
• Changelog for v2026.4.12 / v2026.4.14 / v2026.4.15 — no mention of these pipelines
• Open PRs (#67329, #68041, #67208, #68586 and ~15 others) on the memory-wiki plugin — all bridge-import bug fixes, none adding extractors
• Related roadmap issues:
  • #11946 (Post-Conversation Hook for Automated Knowledge Extraction) — open
  • #62703 (Post-turn capture hook with structured output routing) — open, not merged

Question

1. Are the `concepts/` / `entities/` / `syntheses/` folders intended to be user-maintained editorial destinations (I fill them manually), or are they placeholders for a future auto-generation pipeline (extractor that walks `sources/` and writes these)?

2. If the latter, is there a roadmap / target version for when these pipelines ship? Is it gated on the post-turn hook primitive (#62703) landing first?

3. If there's no plan, would you accept a community-contributed extractor skill that populates these folders based on the current `sources/` corpus (using LLM extraction, written into the existing schema with `<!-- openclaw:wiki:... -->` markers)?

Just trying to understand whether I should (a) wait, (b) build my own, or (c) treat these as manual editorial folders.

Thanks for the great plugin btw — the `sources/` + `reports/` pipelines have been very useful.