agentleFS
sschemelsschemel/second-brain-starter/.agents/skills/sod/SKILL.md

sod

Skill0 starsChanged 16 days agoMITView on GitHub

Start of Day routine for the Second Brain vault. Runs a catch-up harvest if the last /eod was missed, resumes paused context, builds an Eisenhower-matrix briefing from #p1/#p2/overdue dashboard tasks, and sets up today's daily note in 01 - Daily Chats/. Use when the user types /sod, asks to start their day or for a morning briefing of their vault tasks, or the daily-routine hook says /sod hasn't run today.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run date +%F for TODAY and date +%A for the weekday.

    line 13
  2. Very specific:

    Read 99 - System/memory/Working_Context.md.

    line 23

≈ “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

SOD (Start of Day)

Bootstrap the day. Optional focus from the user: $ARGUMENTS

1. Establish dates

Run date +%F for TODAY and date +%A for the weekday. Never guess the date.

1b. Catch-up harvest (only when the last /eod was missed)

Run this step when the daily-routine hook says a catch-up is owed, or when any of these is true: 00 - Inbox/Learnings.md has entries below its heading, the most recent earlier daily note has no ## 🌙 EOD Summary, or backup is enabled (99 - System/memory/backup.md) and the last vault commit (git log -1 --format=%cr) is more than 36 hours old (the daily-routine hook uses the same threshold). Do these parts of the eod skill (.agents/skills/eod/SKILL.md), and nothing else:

  • §3 Learnings harvest (Learnings.md entries only, plus the project-learnings and harness-memory scans). Leave Ideas_Backlog.md and Raw_Captures.md for the next real /eod.
  • §5 Backup, with its secret gate and private-repo check, if backup is enabled. Use the commit message SOD TODAY: catch-up harvest.
  • Skip the summary and the rollover question. Overdue tasks surface in the matrix below instead, and the missed day's note gets no retroactive summary. Report it in one line of the briefing ("Caught up: N learnings → AGENTS.md §7, M → project files, backup <hash or skipped>").

2. Resume context

  • Read 99 - System/memory/Working_Context.md. If it holds a /pause checkpoint, it is the first thing in the briefing ("Resume: …").
  • Find the most recent daily note in 01 - Daily Chats/ dated before TODAY (it may not be yesterday). Read its ## 🌙 EOD Summary and ## 🔭 Tomorrow sections for continuity. If none exist, say so and continue.

3. Discover tasks

Tasks live in each project's _Dashboard.md (SSOT), plus 00 - Inbox/ and 20 - Areas/. Search only those. Specs and plans carry acceptance-criteria checkboxes, often tagged and dated, and they would flood the matrix:

grep -n -H -e '- \[ \]' "10 - Projects"/*/_Dashboard.md "00 - Inbox"/*.md 2>/dev/null; grep -rn --include='*.md' -e '- \[ \]' "20 - Areas" 2>/dev/null

Drop tasks that sit under a heading containing "Archived". Keep a task if any of these is true: tagged #p1 or #p2, has a 📅 date on or before TODAY (overdue or due today), or is tagged @waiting. Note each task's project (from its folder) and file:line. Compute "overdue" from each task's own 📅 date against TODAY, never from an earlier note's wording.

Stale check: a task is a stale suspect when it is more than 7 days overdue, or when it is #p1 and undated but its project's recent notes (Working_Context.md, the dashboard's newest Staging lines) show the work moved on (merged, superseded, replaced). Collect these into the 🧹 list rather than the 🔥 quadrant.

4. Build the Eisenhower matrix

QuadrantRule
🔥 Do now (urgent + important)#p1, or any task that is overdue or due today
📅 Schedule (important, not urgent)#p2 with a future or missing date
⏳ Waiting / Delegate@waiting (show how long it has been waiting when the date allows)
🧹 Later / staleoverdue #p3, untagged overdue tasks that aren't clearly important, and the stale suspects from step 3. Show a count per project and the 3–5 most suspicious, and offer one batch cleanup (close, re-date, or move under the dashboard's Archived heading).

If a focus was given, filter to it. Cap 🔥 Do now at about 5 items, ranked by what is due soonest and what the active checkpoints in Working_Context.md are blocked on. If more qualify, say how many and that the rest are in the dashboards. Do not modify task lines during SOD unless the user accepts the stale-cleanup offer; rescheduling otherwise happens at /eod or on request.

5. Write today's note

Target: 01 - Daily Chats/TODAY.md.

  • If it doesn't exist, create it:
    # TODAY (Weekday)
    
    ## 🎯 Priorities
    <matrix from step 4, each item as "task text — [[Project/_Dashboard]]">
    
    ## 💬 Sessions
    
    ## 🌙 EOD Summary
  • If it exists, replace only the ## 🎯 Priorities section and leave everything else untouched. If the note has no section headings yet (sessions often append bare bullets), put ## 🎯 Priorities under the title and a ## 💬 Sessions heading above the existing bullets.
  • Reference tasks rather than copying them as new checkboxes (SSOT): write plain bullets, not - [ ] lines, so the Tasks plugin doesn't double-count them.

6. Brief the user

In chat, give a crisp briefing: the catch-up line and the resume line (if any), the matrix, and one recommended first move. Keep it scannable, around 15 lines.

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.