agamaroraagamarora/second-brain/.claude/skills/deep-sync/SKILL.md

deep-sync

Skill2 starsChanged 7 months agoMITView on GitHub

Weekly session-end synthesis. Everything in /quick-sync plus drift detection, wiki snapshot via /publish-context, and retrospective notes. Run once a week (Friday recommended).

Commits and pushes

In this repository12 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- absorbSkill
- captureSkill
- extract-entitiesSkill
- helpSkill
- ingestSkill
- lintSkill
- new-pageSkill
- orientSkill
- publish-contextSkill
- querySkill
- quick-syncSkill

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
Read frontmatter + ## Sources that shaped this page table
line 18
Very specific:
If 5+ unabsorbed decisions for a single page → flag as DRIFT
line 20
Distinctive:
If drift list is non-empty and ≤3 pages → enrich each one inline before finishing.
line 30
Distinctive:
If drift list is >3 pages → enrich the top 2 by unabsorbed count.
line 32
Very specific:
Call /lint (report-only, no --fix).
line 36
Very specific:
Call /publish-context to export a dated snapshot of wiki/ + journal/NOW.md + decisions/ to archive/context-export/<YYYY-MM-DD>/ for …
line 40

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

SKILL.md

Weekly deep synthesis. Run after /quick-sync once per week.

Steps

1. Run /quick-sync

Delegate to /quick-sync first. Inbox processed, decisions logged, pages enriched, journal updated. Do NOT push yet — this skill will push at the end.

2. Drift detection

For every page in wiki/ and every file in decisions/:

Read frontmatter + ## Sources that shaped this page table
Compute: how many decisions reference this page but this page was not enriched in the last 30 days?
If 5+ unabsorbed decisions for a single page → flag as DRIFT

Output a drift list:

Drift check:
- wiki/concepts/<slug>.md — N unabsorbed decisions (most recent: <date>)
- ...

If drift list is non-empty and ≤3 pages → enrich each one inline before finishing. Append a new dated section per page synthesizing the drifting decisions.

If drift list is >3 pages → enrich the top 2 by unabsorbed count. Log the rest for next session.

3. Lint the brain

Call /lint (report-only, no --fix). Include the findings summary in the retro note (step 4). High-value findings — broken links, contradictions — should surface into next week's tracks rather than being silently filed.

4. Wiki snapshot

Call /publish-context to export a dated snapshot of wiki/ + journal/NOW.md + decisions/ to archive/context-export/<YYYY-MM-DD>/ for cross-session portability.

5. Retrospective note

Append to journal/retro-<YYYY-ww>.md (ISO week number):

# Retrospective: week <ww> of <YYYY>

## What shipped
<list, drawn from journal/DONE.md entries this week>

## Decisions this week
<list, from decisions/ filenames matching this week's dates>

## Drift detected
<list from step 2>

## Lint findings
<summary from step 3 — orphans, broken links, contradictions>

## Pages enriched
<list>

## What to do next week
<1-3 items>

6. Append to log.md

## [YYYY-MM-DD] deep-sync
- Drift pages: N
- Lint findings: N (O orphans, B broken links, S stale, C contradictions)
- Snapshot: archive/context-export/<date>/
- Retro: journal/retro-<YYYY-ww>.md

7. Commit and push

git add -A
git commit -m "deep-sync: week <ww>"
git push

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

It worked
Partly
It didn't

What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free