Hot Cache Vault Layer¶

The hot cache is an optional wiki/hot.md rolling summary for workspace orientation.

Read order:

1. wiki/hot.md when present and enabled.
2. wiki/index.md.
3. Recent wiki/log.md.
4. Target wiki pages.

API¶

• GET /api/workspaces/{id}/hot-cache
• PUT /api/workspaces/{id}/hot-cache/config
• POST /api/workspaces/{id}/hot-cache/refresh

Refresh is deterministic without an API key: callers can pass summary or recent_text, and Keprix caps the rendered hot cache to roughly 500 words.