agentleFS
cacardinalcacardinal/laila/.claude/skills/quarterly-refresh/SKILL.md

quarterly-refresh

Skill1 starsChanged 4 months agoMITView on GitHub

Interactive quarter-boundary ritual — grade the outgoing quarter's goal tree against its KPIs with evidence, harvest lessons, archive it honestly, and propose next quarter's slate. Use when Alex says "quarterly planning", "quarterly refresh", "new quarter", "Q1/Q2/Q3/Q4 planning", "grade the quarter", or a quarter boundary is near (an optional scheduled nudge may fire ~7 days out). Operationalizes the Quarterly Planning Specifics section of templates/planning-session.md. NEVER headless — interactive only; writes NOTHING until Alex approves.

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 7 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    NOT for: weekly planning, "capacity planning" (both stay in templates/planning-session.md), monthly KR check-ins, or /os-audit (system …

    line 20
  2. Very specific:

    Run templates/planning-session.md Steps 1-4 (state review, domain scan, external sync, capacity table).

    line 31
  3. Distinctive:

    Before anything is archived, ask (one at a time, conversationally):

    line 53
  4. Distinctive:

    Only after Alex approves the full grade table:

    line 65
  5. Distinctive:

    Write the new tree to the goal source of truth.

    line 87
  6. Very specific:

    If priorities shifted, propose the state/strategy.md edit as a diff and wait for a separate yes — strategy prose is never rewritten without …

    line 89
  7. Very specific:

    Write the approved knowledge/tacit/ and decision-log entries.

    line 91
  8. Very specific:

    If you want a reminder, add a scheduled job (launchd/cron) that fires ~7 days before quarter end and sends one line to the command channel: …

    line 106
  9. Distinctive:

    Interactive only, ever: If running headless (no human turn-taking), STOP and exit.

    line 116
  10. Distinctive:

    Before declaring this skill done (and again after edits):

    line 128
  11. Very specific:

    Export test: after an approved run, confirm the source of truth and state/goals.md agree, and the outgoing quarter appears archived with …

    line 137
  12. Distinctive:

    After each run: if Alex corrected a grade classification, an evidence citation, or the slate format, this skill is wrong, not Alex.

    line 142
  13. Distinctive:

    If the same correction happens twice, the fix is MANDATORY before the next run.

    line 146

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Quarterly Refresh

Close the quarter with evidence, not vibes, then open the next one with a slate Alex actually has capacity for. Five phases: Grade, Harvest, Archive, Set, Wire the cadence.

NEVER HEADLESS. This skill runs only in an interactive session with Alex at the keyboard. It writes NOTHING — no goal edits, no archives, no export regeneration — before Alex approves the full proposed slate. A scheduled job may nudge that a quarter boundary is near (see Timing), but the nudge only notifies; it never runs this skill.

When to use

  • "quarterly planning", "quarterly refresh", "new quarter", "Q1/Q2/Q3/Q4 planning", "grade the quarter"
  • At a quarter boundary, or when the quarter-end nudge fires
  • NOT for: weekly planning, "capacity planning" (both stay in templates/planning-session.md), monthly KR check-ins, or /os-audit (system health, not goals)

This skill fills the "Quarterly Planning Specifics" slot in templates/planning-session.md — run it instead of freehanding that section. For everything upstream of the quarterly specifics (state review, domain scan, external sync, capacity table), follow that template's Steps 1-4; don't duplicate them here.

Steps

Phase 0 — Context (read-only)

Run templates/planning-session.md Steps 1-4 (state review, domain scan, external sync, capacity table). Then pull the outgoing quarter's full goal tree from the goal source of truth — the CRM's goal objects if one is wired, or state/goals.md directly if not. Per the cross-domain rules, state/goals.md is a generated export when a CRM is wired; grade against the source, never the export.

Phase 1 — Grade

For each Objective in the outgoing tree:

  1. Gather evidence per KPI/KR — the tracking file row, the shipped artifact, the number. Cite it (domains/career/tracking/status.md, the published post count, "5 of 8 calls"). No evidence found = say so; never invent a status.
  2. Score each KR 0.0-1.0 per templates/okr-architecture.md (0.7+ is success), then classify the Objective: HIT (KRs ≥0.7), PARTIAL (real progress, short), MISS (little progress), ABANDONED (deliberately dropped mid-quarter).
  3. One-line "why" for anything not HIT.
  4. Present as a table and walk it Objective by Objective. Alex confirms or corrects each grade — the grades are Alex's; the evidence-gathering is Laila's. Record corrections.

Phase 2 — Harvest

Before anything is archived, ask (one at a time, conversationally):

  • For each MISS/ABANDONED: wrong goal or wrong execution? Wrong goals die; wrong execution may deserve another attempt with a changed approach.
  • Which PARTIALs/MISSes roll forward into next quarter vs. die here? Rolling forward is a deliberate choice, not a default — a goal that rolls twice is a wrong goal.
  • Any lessons worth keeping? Propose candidate entries for knowledge/tacit/ and a decision-log line for knowledge/decisions/YYYY-MM.md — propose only; these get written with everything else at Phase 4 approval, never auto-written.

Phase 3 — Archive (first write gate)

Only after Alex approves the full grade table:

  1. Mark the outgoing quarter's goals archived/closed in the source of truth, carrying the grade and the one-line "why".
  2. Regenerate the state/goals.md export (or, if goals.md IS the source, move the graded quarter to a "Completed" section per templates/planning-session.md).

Phase 4 — Set

Propose the incoming quarter's slate, shaped by templates/okr-architecture.md:

  • 3-5 Objectives max — qualitative, memorable, verb-first. Cross-check against state/strategy.md focus areas and explicit non-priorities; flag overcommitment out loud (a 6th Objective needs one of the 5 to die, and an Objective outside the strategy's priorities needs a strategy conversation, not a quiet insertion).
  • Each Objective: 1-3 Initiatives and measurable KPIs/KRs (outcome, number, date).
  • Fold in the Phase 2 roll-forwards, reshaped — never pasted in with the old wording.

Present the FULL slate as ONE proposal — Objectives, Initiatives, KPIs, plus the proposed tacit/decision-log entries from Phase 2. Iterate until Alex approves the whole thing; no partial writes mid-iteration. On explicit approval:

  1. Write the new tree to the goal source of truth.
  2. Regenerate the state/goals.md export.
  3. If priorities shifted, propose the state/strategy.md edit as a diff and wait for a separate yes — strategy prose is never rewritten without explicit approval.
  4. Write the approved knowledge/tacit/ and decision-log entries.

Phase 5 — Wire the cadence

Close the loop so the new tree actually gets used:

  • Remind Alex that weekly planning and the daily brief now key off the new tree — the next "Weekly planning" session should pull priorities from it.
  • Suggest (Tier 3, propose first) a mid-quarter check row in state/active-tasks.json (owner: laila, due ~week 6-7: "mid-quarter KR check-in — score each KR 0.0-1.0").
  • Suggest domain tracking/status.md updates where an Objective changed a domain's projects — per templates/planning-session.md Step 6, also proposed, not silent.

Timing — the optional nudge (scheduled-job pattern)

If you want a reminder, add a scheduled job (launchd/cron) that fires ~7 days before quarter end and sends one line to the command channel: "Q ends in 7 days — say 'quarterly refresh' when you're ready." Use the headless AGENT_RUN pattern from docs/headless-sessions.md for the notification only, then install the job and regenerate the registry (see docs/background-monitoring.md — state/loops-registry.json is generated, never hand-edited). The nudge NEVER invokes this skill — grading a quarter is a conversation, not a batch job.

Rules

  • Interactive only, ever. If running headless (no human turn-taking), STOP and exit.
  • Two approval gates, zero early writes: nothing before the grade table is approved (gate 1 → archive), nothing else before the full slate is approved (gate 2 → set). Phases 0-2 are strictly read-only.
  • Grades are Alex's; evidence is Laila's. Never argue a grade — record the correction.
  • Evidence or admit the gap. "Felt productive" is not a KR score.
  • 3-5 Objectives is a ceiling, not a target. Push back on the 6th.
  • state/goals.md is touched only as a regenerated export unless it IS the source.
  • state/strategy.md edits are always a separately-approved diff.

Verification (MANDATORY)

Before declaring this skill done (and again after edits):

  • Cold test: fresh session, "quarterly refresh". It must run planning-session Steps 1-4, produce an evidence-cited grade table, and reach the Alex-confirmation loop with zero writes having occurred.
  • Gate test: decline the grade table — verify nothing was archived and no file changed. Approve grades but reject the slate twice — verify iteration happens with still zero new-quarter writes until the explicit yes.
  • Overcommit test: propose a 6-Objective slate as Alex — the skill must flag it and ask which one dies, not silently accept.
  • Export test: after an approved run, confirm the source of truth and state/goals.md agree, and the outgoing quarter appears archived with grades attached.

Self-improvement (MANDATORY)

  • After each run: if Alex corrected a grade classification, an evidence citation, or the slate format, this skill is wrong, not Alex. Propose the edit to this file immediately, or log {"skill": "quarterly-refresh", "correction": "...", "date": "YYYY-MM-DD"} to state/skill-feedback.jsonl.
  • If the same correction happens twice, the fix is MANDATORY before the next run.

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.