second-brain-review

1# Review the vault
2 
3Capture systems fail at the point where nothing comes back out. A review is the
4loop that closes: it turns a week of ingestion into something the owner reads
5and acts on, and it is what makes the vault feel alive rather than archival.
6 
7## Core rule
8 
9A review reports on the vault, not on the agent's activity. Nobody needs to
10know how many tool calls ran; they need to know what they now know.
11 
12## Workflow
13 
141. **Diff the period.** New and updated pages since the last review, from
15   `log.md` and frontmatter dates.
162. **Find the clusters.** Which concepts gained the most links. That is where
17   the owner's attention actually went, which is often not where they think it
18   went.
193. **Surface contradictions** recorded during the period and still unresolved.
204. **List open questions** from concept and synthesis pages.
215. **Find the neglected.** Pages linked often but thin, or gaps in `index.md`
22   that keep getting referenced.
236. **Recommend three things to read or write next**, each tied to a specific
24   page.
25 
26## Output format
27 
28```
29## Week of <date>
30 
31Added: <n> pages, <n> sources ingested
32 
33### Where your attention went
34<two or three sentences on the clusters>
35 
36### Unresolved
37<contradictions and open questions, each linked>
38 
39### Thin spots
40<pages that are referenced more than they deserve given their content>
41 
42### Next
431. <specific action tied to a page>
442.
453.
46```
47 
48## Calibration
49 
50Reviews that list everything get skimmed and then ignored. Keep it under a
51…