agentleFS
eugenezhangco-speceugenezhangco-spec/athena/.claude/skills/debrief/SKILL.md

debrief

Skill0 starsChanged 7 months agoMITView on GitHub

End-of-day check-in. Fixed 5-step sequence: (1) list today's todos and ask what was finished, reallocate the rest to tomorrow; (2) pull today's emails 8am-6pm grouped urgent / important / noise, draft 2-sentence replies for urgent items; (3) show tomorrow's calendar; (4) ask for thoughts on today, coach, and log to diary/patterns. Use when the user says 'debrief', 'wrap up the day', 'end of day', 'how did today go', 'close out the day', 'night', or 'goodnight'. Do NOT use for full daily planning (use day-planner). Do NOT use for weekly review (use weekly-review).

31 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use brief, review and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • AnthropicSkill

    brief

    Generate contextual briefings for legal work — daily summary, topic research, or incident response. Use when starting your day and need a scan of legal-relevant items across email, calendar, and contracts, when researching a specific legal question across internal sources, or when a developing situation (data breach, litigation threat, regulatory inquiry) needs rapid context.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 3 projects that use this

  • MicrosoftSkill

    handoff

    Compact the current conversation into a handoff document for another agent to pick up.

    Also in 3 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. 0 are common, 14 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 28 edited versions we read keep these.

  • 14

    Distinctive

    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. Distinctive:

    Read these files before starting:

    line 27
  2. Distinctive:

    Before saying anything, gather all data in parallel:

    line 47
  3. Distinctive:

    Wait for the user's reply.

    line 68
  4. Distinctive:

    If they finished something not on the list, capture it.

    line 73
  5. Distinctive:

    Wait for the user to tell you which replies to send, edit, or skip.

    line 96
  6. Very specific:

    If a pattern was confirmed or a new one observed, update patterns.md using the pattern lifecycle rules (WATCHING → ACTIVE → RESOLVED).

    line 124
  7. Distinctive:

    Create new tasks the user mentioned during Step 4.

    line 147
  8. Very specific:

    Write a "Carried forward" block into tomorrow's section of day-ledger.md.

    line 150
  9. Distinctive:

    Not a performance review. The assistant is not grading the user's day.

    line 181
  10. Distinctive:

    Not a full backlog audit. That is the weekly review.

    line 182
  11. Distinctive:

    Not a planning session. That is the day planner.

    line 183
  12. Distinctive:

    Not a guilt trip. If the day went sideways, acknowledge it and move on.

    line 184
  13. Distinctive:

    Not optional forever. But it should feel easy enough that the user wants to do it.

    line 185
  14. Distinctive:

    If the day was rough: "Yeah, that is a lot.

    line 227
  15. Distinctive:

    If the day was great: "That is a win.

    line 228
  16. Distinctive:

    Always close clean. No trailing offers to help.

    line 229

“Common” is measured: kept by at least 6 in 10 of the 28 edited versions of this file we could read. “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

Skill: Debrief

Trigger: "debrief", "end of day", "wrap up the day", "goodnight", "night"


What This Skill Does

Closes the daily loop. Fixed 5-step flow every time:

  1. Today's todos — list them, user says what's done, rest carries to tomorrow.
  2. Today's emails (8am–6pm) — grouped urgent / important / noise, with 2-sentence draft replies for urgent items.
  3. Tomorrow's calendar — clean chronological view with flags.
  4. Thoughts on today — user reflects, assistant coaches, logs to diary + patterns.
  5. Commit — write everything to ledger, snapshot, task database, patterns, MemPalace.

The assistant leads. The user answers. No mode branching.


Dependencies (Load Before Running)

Read these files before starting:

  1. personal/day-ledger.md -- TODAY plan section (if it exists). Primary reference for today's todos/reminders.
  2. personal/snapshot.md -- current flags and session context.
  3. personal/patterns.md -- behavioral patterns to check against today.
  4. personal/goals.md -- quarterly targets for connecting today to the bigger picture.
  5. .claude/rules/evolution.md -- post-session self-check protocol.

Query these external sources (silent, before any response):

  • Google Calendar: today's actual events AND tomorrow's events across all calendars.
  • Gmail: all emails received today between 08:00 and 18:00 in user's timezone.
  • Task database / Notion: tasks and reminders due today.

The Debrief Flow

Fixed 5-step sequence. The assistant leads each step, waits for the user's response, then moves to the next.

Phase 0: Silent Data Gathering

Before saying anything, gather all data in parallel:

  1. Today's todos/reminders — from day-ledger.md TODAY Plan section, plus any reminders in Notion or the task database due today.
  2. Today's emails — Gmail search for messages received today between 08:00 and 18:00 user local time. Pull sender, subject, snippet.
  3. Tomorrow's calendar — Google Calendar, all calendars, for tomorrow's date in user's timezone.
  4. Active patterns from patterns.md.

The assistant now has everything. No need to ask the user to supply any of it.

Step 1: Today's Todos

Present every todo/reminder that was on today's list. Clean numbered list. No commentary yet.

"Here is what was on the list for today:

  1. [Task A]
  2. [Task B]
  3. [Reminder C] ...

What did you finish? Anything I should know about the rest?"

Wait for the user's reply.

When they respond:

  • Mark finished items as Done (task database + ledger).
  • Everything they did not finish gets carried to tomorrow — added to tomorrow's Plan section in day-ledger.md and, if applicable, rescheduled in the task database. Note the reason if the user gave one.
  • If they finished something not on the list, capture it.

Step 2: Today's Emails (08:00 – 18:00)

Present emails grouped into three buckets. Keep it tight — sender + one-line subject/summary per email.

URGENT — reply today
- [Sender]: [subject] — [one-line why it's urgent]
  Draft reply:
  "[2-sentence draft the user can send with minor edits]"

IMPORTANT — not urgent
- [Sender]: [subject] — [one-line summary]

NEWSLETTERS / NOISE
- [count] items, skim or archive: [brief list or categories]

Urgency rule: "urgent" means something the sender is waiting on today or there's a real deadline. Not just that it feels important. Be strict. If unsure, put it in Important.

Drafts: match the user's voice-dna.md. Two sentences. No filler. No "Hope you're well." No emojis. No em dashes.

Wait for the user to tell you which replies to send, edit, or skip.

Step 3: Tomorrow's Calendar

Show tomorrow's calendar in clean chronological order. Day of week + date at the top.

TOMORROW — [Day, Date]
08:30  [Event] — [location/attendees if relevant]
10:00  [Event]
14:00  [Event]
...

Flag anything worth flagging: back-to-back meetings, deep work block at risk, travel time gaps, a meeting with no prep.

Step 4: Thoughts on Today

Open the floor.

"Any thoughts on today? What was heavy, what felt good, what is sitting with you."

Let the user talk. Do not interrupt with advice. Listen first.

When they finish:

  • Pick the one or two most important threads.
  • Coach, not counsel. Reference patterns.md if a pattern showed up. Reference goals.md if today drifted from priorities.
  • Log the reflection to personal/diary.md (create if it does not exist) as a dated entry — their words first, then the assistant's observation.
  • If a pattern was confirmed or a new one observed, update patterns.md using the pattern lifecycle rules (WATCHING → ACTIVE → RESOLVED).
  • Mine any user-stated facts, preferences, or decisions to MemPalace with source tag debrief.

Then: "Locking this in." (Commit point -- see Phase 3.)


Phase 3: The Commit

At the end of Step 4, execute all writes in one pass:

1. Day Ledger

Append the Debrief section to today's entry in personal/day-ledger.md:

  • Committed at (current time, user's timezone)
  • Finished today: [tasks the user confirmed done]
  • Rolled to tomorrow: [tasks moved forward, with reason if given]
  • Emails handled: [count urgent / important / noise; which drafts were sent or edited]
  • User's thoughts: (their words, compressed but faithful)
  • Assistant observed: (patterns, dots connected, one coaching line)

2. Task Database

  • Mark finished tasks as Done.
  • Reschedule unfinished tasks to tomorrow's date (dedup before creating new ones).
  • Create new tasks the user mentioned during Step 4.

3. Tomorrow's Plan Seed

Write a "Carried forward" block into tomorrow's section of day-ledger.md. This is what the morning briefing and day planner read first.

4. Snapshot

Update personal/snapshot.md:

  • Last Session: today's date, debrief completed.
  • Flags: add/resolve based on the conversation.

5. Patterns + Diary

  • personal/patterns.md: confirm, promote, or resolve patterns per lifecycle rules.
  • personal/diary.md: dated entry with user's reflection + assistant's observation. Create file if it does not exist.

6. MemPalace

Mine user-stated facts, preferences, or decisions from Step 4 to MemPalace with source tag debrief.

7. Sign-off

One clean block:

DEBRIEF LOCKED

Finished: [N tasks]
Rolled to tomorrow: [N tasks]
Emails: [X urgent replied, Y important flagged, Z archived]
Tomorrow's first move: [specific task]
[One human observation — not mechanical]

What the Debrief is NOT

  • Not a performance review. The assistant is not grading the user's day.
  • Not a full backlog audit. That is the weekly review.
  • Not a planning session. That is the day planner.
  • Not a guilt trip. If the day went sideways, acknowledge it and move on. "Rough day. Tomorrow is a reset."
  • Not optional forever. But it should feel easy enough that the user wants to do it. A few minutes, every night.

How This Feeds the System

Morning Briefing
  reads: yesterday debrief "Carried forward" from ledger
  reads: snapshot.md flags
  -> sends compressed wake-up message

Day Planner
  reads: today ledger (if briefing seeded anything)
  reads: snapshot.md, patterns.md, goals.md, task database, calendar
  -> writes Plan section to ledger on commit

The Day (user lives their life)

Debrief
  reads: today Plan from ledger
  reads: calendar (actual), task database (actual), patterns.md
  -> writes Debrief section to ledger on commit
  -> updates snapshot.md, patterns.md, task database
  -> seeds tomorrow "Carried forward"

Weekly Review
  reads: 7 days of ledger entries (plan + debrief pairs)
  -> full audit, pattern confirmation, goal tracking, architecture hygiene

The daily loop feeds the weekly loop. The weekly loop feeds the quarterly goals. Nothing is reconstructed from scratch. Everything compounds.


Tone in Debrief

End of day. The user is tired. Match the energy.

  • Warmer than morning. Less push, more presence.
  • Efficient, no fluff, but not cold. "Good day. Three things moved. One thing to carry."
  • During Step 4 (Thoughts on today): slow down. Let silences land. This is where the real stuff surfaces.
  • If the day was rough: "Yeah, that is a lot. Tomorrow is a different day." Not "I am sorry to hear that."
  • If the day was great: "That is a win. Remember this feeling when the next hard day hits."
  • Always close clean. No trailing offers to help. The debrief closes the day. Let it close.

Context to Load

  • personal/day-ledger.md -- FIRST. Today plan section.
  • personal/snapshot.md -- Current flags, last session.
  • personal/patterns.md -- Active behavioral patterns.
  • personal/goals.md -- Quarterly targets (for connecting today to the bigger picture).
  • .claude/rules/evolution.md -- Self-check protocol.
  • .claude/rules/operating-rules.md -- Already loaded. Timezone, day type, priority stack.

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.