goaly-coaching-prep

name	goaly-coaching-prep
description	Use when preparing for coaching with [Coach], collecting KPIs, or building a coaching agenda. Triggers: 'prep for [Coach]', 'coaching prep', 'coaching agenda'.
allowed-tools	["Read","Grep","Glob","Bash","Edit","Write","AskUserQuestion"]

Before You Start

1. Read .claude/skills/_shared/conventions.md
2. Read gotchas.md in this directory

Learned Preferences

!tail -20 "${CLAUDE_SKILL_DIR}/data/feedback.log" 2>/dev/null || echo "No feedback yet — preferences will accumulate over sessions."

/goaly-coaching-prep — Coaching Session Prep

Prepares a structured agenda for [Owner]'s bi-weekly coaching session with [Coach]. Collects KPI updates, reviews progress since last session, surfaces standing items, and flags patterns.

Context

• Today: !date +%Y-%m-%d
• Day: !date +%A
• Uncommitted changes: !git status --short notion-mirror/

---

STEP 1: Freshness Check

Run sync verification per _shared/conventions.md. If uncommitted notion-mirror/ changes exist, warn [Owner] before proceeding.

---

STEP 2: Parallel Data Collection

Collect ALL data in a single parallel phase. Launch every group simultaneously.

Group A — qmd Semantic Queries

Run in parallel:

qmd query "[Coach] coaching action items commitments next steps"
qmd query "coaching progress goals KPIs blockers"
qmd query "deprioritized parked deferred coaching"

Group B — Grep Frontmatter Queries

Run ALL in parallel:

Grep "^lifecycle: Active" notion-mirror/goals/
Grep "^lifecycle: Active" notion-mirror/kpis/
Grep "^lifecycle: Active" notion-mirror/projects/
Grep "^status: Done This Week" notion-mirror/tasks/
Grep "^status: In progress" notion-mirror/tasks/
Grep "^status: Planned this week" notion-mirror/tasks/
Grep "^status: Deprioritized" notion-mirror/tasks/
Grep "## Coaching Accountability" notion-mirror/tasks/

Group C — File Reads

Run in parallel:

Read ~/.claude/projects/-Users-dan-Admin-Goals---Tasks/memory/MEMORY.md
Glob "granola-mirror/meetings/*coaching*" OR Glob "granola-mirror/meetings/*[coach]*"

Group D — Read All Matched Files

After Groups B and C return, read ALL matched files in parallel:

• Every file from active goals, KPIs, projects
• Every file from done/in-progress/planned/deprioritized tasks
• Most recent Granola coaching transcript (if any)

For KPI files, extract: current_value, target_value, _progress, _gap, unit, goal.

Fallback: Granola MCP

Only if qmd + local granola-mirror return no recent coaching sessions, use query_granola_meetings for "[Coach] coaching". If unavailable, proceed with local data and inform [Owner].

---

Data collection complete. Steps 3-8 use collected data only — no new I/O except KPI file edits and git commands.

---

Coaching Accountability Surface

Uses: Grep results for "## Coaching Accountability" in notion-mirror/tasks/. No new I/O.

For each task with a "Coaching Accountability" section in its body:

1. Read the task file
2. Extract: title, status, sessions-committed count, dates committed, frog type, first micro-step
3. Calculate days since first commitment date

Present as table:

Task	Status	Sessions	First Committed	Days Stalled	Frog Type

Rules:

• Sort by sessions-committed DESC (most repeated commitments first)
• Flag any task committed 3+ sessions as "RECURRING STALL — discuss pattern, not just task"
• Include in [Coach] agenda as "Accountability Check" section
• If a task was completed since last coaching session, celebrate it: "FROG EATEN: [task]"

---

STEP 3: Flag Unmeasured Goals & Deprioritized Items

1. Unmeasured goals: For each active goal, check if any active KPI has a matching goal_id. Flag goals with zero KPIs.
2. Deprioritized items: List deprioritized tasks (confirm still parked). Cross-reference with qmd:deprioritized.
3. Stalled tasks: From in-progress and planned tasks, flag any where _notion_edited > 14 days ago.

---

STEP 4: Collect KPI Values

Use AskUserQuestion with structured choices for each Active KPI:

KPI	Current	Target	Progress	Gap
MRR	EUR X	EUR Y	Z%	EUR N

MRR is a standing question every session (see gotchas.md).

---

STEP 5: Update KPI Files

Edit each KPI file where [Owner] provided an updated value — change current_value: only. Do NOT touch read-only _ fields.

Commit KPI updates separately:

git add notion-mirror/kpis/
git commit -m "chore: update KPI values for coaching prep" </dev/null 2>&1

---

STEP 6: Standing Items

Extract "Coaching Prep -- Standing Items" from MEMORY.md (collected in Step 2). Present as checklist:

### Standing Items
- [ ] Current MRR: EUR X (confirmed in Step 4)
- [ ] [each standing item from MEMORY.md, verbatim]

Do NOT hardcode standing items — always read them fresh from MEMORY.md.

---

STEP 7: Build Agenda

## Coaching Agenda — [Coach] ([date])

### Progress Since Last Session
- [completed tasks, KPIs that moved, KPIs that didn't]

### Current Focus & Blockers
- [In progress and Planned tasks, stalled tasks from Step 3]

### Metrics Snapshot
| KPI | Current | Target | Progress | Trend |
|-----|---------|--------|----------|-------|

### Accountability Check
- [from Coaching Accountability Surface — frog table, recurring stalls, frogs eaten]

### Standing Items
- [ ] [from Step 6]

### Patterns & Flags
- [unmeasured goals, initiation avoidance, deprioritized items]

### Questions for [Coach]
- [open threads from qmd:coaching-actions, unresolved items]

---

STEP 8: Present & Close

Present the compiled prep to [Owner]. End with: "After the call, run /goaly-review-meeting coaching to process the transcript."

Commit remaining changes and verify sync per _shared/conventions.md. Check memory enrichment: coaching insights to dan-profile.md, pipeline changes to revenue-strategy.md, new standing items to MEMORY.md.

---

Session Learning

When [Owner] corrects your output or expresses a preference during this session, immediately append it to the feedback log:

mkdir -p "${CLAUDE_SKILL_DIR}/data" && echo "$(date +%Y-%m-%d) <preference description>" >> "${CLAUDE_SKILL_DIR}/data/feedback.log"

Only log general preferences that apply to future sessions — skip task-specific corrections. Use judgment on detail: some preferences are one line, others need a sentence or two of context to be useful in future sessions.

---

Error Handling

Failure	Fallback
qmd returns zero results	Warn [Owner], suggest qmd update && qmd embed, fall back to Grep
Granola MCP unavailable	Proceed with local granola-mirror data only
No prior coaching interactions	Build agenda from KPI/task state only

---

Related Files

Purpose	Path
Skill gotchas	gotchas.md (this directory)
Shared conventions	.claude/skills/_shared/conventions.md
KPI files	notion-mirror/kpis/
Goal files	notion-mirror/goals/
Task files	notion-mirror/tasks/
Project files	notion-mirror/projects/
Meeting transcripts	granola-mirror/meetings/
MEMORY.md (standing items)	~/.claude/projects/-Users-dan-Admin-Goals---Tasks/memory/MEMORY.md
[Owner] profile	~/.claude/projects/-Users-dan-Admin-Goals---Tasks/memory/dan-profile.md
Revenue strategy	~/.claude/projects/-Users-dan-Admin-Goals---Tasks/memory/revenue-strategy.md
Sync log	.sync-log/push.log