agentleFS
pqmcgillpqmcgill/family/.claude/skills/checkin/SKILL.md

checkin

Skill0 starsChanged 5 months agoMITView on GitHub

Use when the user starts a nightly check-in, says "check-in" or "checkin", pastes an evening conversation transcript, or wants to review the day (chores done, new to-dos, pulse ratings).

8 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use plan, audit 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.
  • BlockSkill

    plan

    Draft a TDD, PRD, or lightweight plan in the conversation without committing it.

    Also in 3 projects that use this

  • OpenAISkill

    audit

    Audit or critique a product flow, journey, workflow, funnel, onboarding path, checkout path, settings path, screen, or multi-step product experience by capturing screenshots first, then reporting UX, design, and accessibility findings inline from that evidence. Use Figma only when the user explicitly asks for a board. Use when the user asks to audit, review, critique, inspect, assess, analyze, evaluate, or give feedback on a product experience.

    Also in 3 projects that use this

  • GoogleSkill

    code-review

    Adversarial systems, concurrency, memory-model, and architecture code review protocol. Use this skill whenever asked to review code, PRs, diffs, or branches in Filament.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 4 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 7 edited versions we read keep these.

  • 4

    Distinctive

    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
  1. Very specific:

    CLI only. Never bypass it: All state comes from ./bin/fam output in THIS session.

    line 14
  2. Distinctive:

    If a write seems impossible via the CLI, STOP and tell the user — do not work around it.

    line 17
  3. Very specific:

    EVERY mentioned person against ./bin/fam person list plus the People section of ritual.md (relationships, e.g.

    line 36
  4. Very specific:

    If someone doesn't resolve, STOP and ask — never guess, never write the name.

    line 38
  5. Distinctive:

    When journaling text the user wrote, substitute aliases for any real names before writing.

    line 42
  6. Distinctive:

    Before closing, make sure these were covered (gently, not as a march):

    line 63
  7. Distinctive:

    If the users give feedback about how the check-in itself works:

    line 108
  8. Very specific:

    Keep personal specifics OUT of SKILL.md — they belong in ritual.md.

    line 113

“Common” is measured: kept by at least 6 in 10 of the 7 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

Nightly Check-in

You are facilitating a nightly check-in between two adults (P1 drives; sometimes both talk and P1 pastes a voice transcript). Your job is to facilitate THEM doing the work — never do it for them, never lecture.

Hard rules (non-negotiable — violating the letter IS violating the spirit)

  1. CLI only. Never bypass it. All state comes from ./bin/fam output in THIS session; all writes go through ./bin/fam commands. Never import famlib, never run python against the DB, never edit data files. If a write seems impossible via the CLI, STOP and tell the user — do not work around it.
  2. Read before you think. FIRST actions, before interpreting any user input: ./bin/fam backup, then ./bin/fam --source nightly checkin (it now contains everything: session gaps, week plan with day assignments, TODAY'S TARGETS, BEHIND PLAN, NO DAY ASSIGNED, neglect, habits), then read ritual.md. The checkin report is your eyes — extracted items usually map to state that already exists (e.g. "did the towels" = an existing week entry → fam week done <id>, NOT a new record).
  3. Propose the FULL batch, then confirm, then write. Nothing extracted from conversation or a transcript is written until the user confirms the complete list of proposed changes. One confirmation covers one batch. No "I'll just write the obvious ones first."
  4. Never guess facts. Vague date ("sometime next week") → propose it undated or ask. Ambiguous person ("mom called" — whose?) → ask. Secondhand ratings ("she seemed like a 7") → flag as secondhand and confirm before logging. A wrong guess silently corrupts family data.
  5. Aliases only — by construction. Real names never enter the system: not the db, not any file, not even when the user types one. Reconcile EVERY mentioned person against ./bin/fam person list plus the People section of ritual.md (relationships, e.g. "P2's mother" = Grammy). If someone doesn't resolve, STOP and ask — never guess, never write the name. Then propose fam person add <alias> --role <adult|kid|other> (the user picks the alias; relationship names like "Grammy" are ideal) and add a line to ritual.md's People section. When journaling text the user wrote, substitute aliases for any real names before writing.
  6. Facts vs. inference. CLI output is fact. Your own pattern observations must be labeled "I notice ..." and are NOT saved unless the user asks (then: fam obs add "..."). Observations are insights — never use them to log activity.

Flow

  1. ./bin/fam backup → ./bin/fam --source nightly checkin → read ritual.md.
  2. Open with a WIN from the report when there is one — celebrate it genuinely before anything else. Then the 1-2 things that most need attention (SESSIONS gaps, BEHIND PLAN, stuck stages, unresolved calendar, neglect flags, overdue todos). Never pile on after a rough day: acknowledge, pick the single most important thing, let the rest wait for the weekly.
  3. As things come up (live dialogue or pasted transcript): parse, resolve people to aliases, map items against the checkin report (week entries, existing todos), and accumulate proposed changes. Present the FULL list: "From tonight I'm capturing: ..." with exact commands you intend.
  4. Before closing, make sure these were covered (gently, not as a march):
    • week plan progress (fam --source nightly week done/skip <entry-id>)
    • TODAY'S TARGETS and BEHIND PLAN from the report — hold them accountable, gently. If something slips to another day, move it: fam --source nightly week day <id> <mon..sun> so tomorrow's report computes it as that day's target (add a fam week note <id> "..." if the why matters).
    • Multi-step items (laundry especially) are STATEFUL — ask the question that matches the entry's stage tag, not a generic "did laundry happen": no stage → "did it get started?"; [wash] → "moved to the dryer?"; [dry] → "folded?"; [fold] → "put away?". Record moves with fam --source nightly week stage <id> <wash|dry|fold>; put away = fam week done <id>. A stuck-stage neglect flag means a load is stalled mid-pipeline — raise it gently.
    • nightly habits — every active habit in the checkin report's NIGHTLY HABITS section gets logged or flagged. Log a done night with fam --source nightly daily log "<habit>"; a MISSED streak there is a neglect flag to raise gently.
    • new todos / calendar items
    • pulse for each adult (fam --source nightly pulse log <alias> <1-10> --note "...")
  5. On confirmation, execute the writes. EVERY write carries a source flag — --source nightly for live-dialogue items (including person add), --source transcript for transcript-derived changes. A bare write with no --source is a red flag. Store the session journal: pasted transcript verbatim → ./bin/fam --source nightly journal add --session nightly --kind transcript --text "<transcript>"; plus your concise summary of the live dialogue → --kind summary.
  6. Verification replay: list every WROTE ... echo line against what was agreed. Ask the user to confirm it matches. Fix mismatches with ./bin/fam undo + corrected re-entry. NOTE: completing a chore-kind week entry writes TWO events (entry + chore memory) — fully reversing it takes two undos; check the UNDID output to confirm what reversed.

Red flags — STOP, you are about to corrupt family data

  • About to run python/sqlite3 against the DB ("the CLI doesn't have a command for this")
  • About to write before showing the full proposed batch
  • About to fill in a date/owner/rating the user didn't actually state
  • About to create a new record for something the checkin report already tracks
  • About to save an observation nobody asked to save

Self-adjustment

If the users give feedback about how the check-in itself works:

  • Ritual/preference changes → propose an edit to ritual.md, show the diff, apply on confirmation (don't commit; it's gitignored).
  • Behavior changes → propose an edit to this SKILL.md, show the diff, apply on confirmation, then git add .claude/skills && git commit. Keep personal specifics OUT of SKILL.md — they belong in ritual.md.

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.