Ceos Scorecard

ceos-scorecard

Manage the weekly Scorecard — the 5-15 numbers that tell you the health of your business at a glance. Define metrics, log weekly values, and spot trends before they become problems.

When to Use

• "Log scorecard" or "enter this week's numbers"
• "Define a new metric" or "add a metric to the scorecard"
• "How are our metrics?" or "scorecard trend"
• "Show the scorecard" or "weekly numbers"
• "Which metrics are off track?"
• Any weekly Scorecard review (typically during L10 meetings)

Context

Finding the CEOS Repository

Search upward from the current directory for the .ceos marker file. This file marks the root of the CEOS repository.

If .ceos is not found, stop and tell the user: "Not in a CEOS repository. Clone your CEOS repo and run setup.sh first."

Sync before use: Once you find the CEOS root, run git -C <ceos_root> pull --ff-only --quiet 2>/dev/null to get the latest data from teammates. If it fails (conflict or offline), continue silently with local data.

Key Files

Week Format

Weeks follow ISO 8601: YYYY-WNN (e.g., 2026-W07). Week 1 is the week containing the first Thursday of the year.

Metric Status

Each metric value is compared against its goal:

• on_track (green) — meets or exceeds the goal threshold
• off_track (red) — below the goal threshold

The thresholds are defined in data/scorecard/metrics.md.

Process

Mode: Define Metrics

Use when setting up the Scorecard or adding new metrics.

Step 1: Read Current Metrics

Read data/scorecard/metrics.md. If the file doesn't exist, tell the user to run setup.sh init first.

Display the current metrics table.

Step 2: Add or Edit a Metric

For each new metric, collect:

1. Metric name — what are we measuring? (e.g., "Weekly Revenue", "New Customers")
2. Owner — one person responsible for this number
3. Goal — the target value (e.g., "$50,000", "10")
4. Frequency — Weekly (most metrics) or Monthly
5. Green threshold — what value means "on track" (e.g., ">= $50,000")
6. Red threshold — what value means "off track" (e.g., "< $50,000")

Step 3: Validate

• 5-15 metrics total. Fewer means blind spots. More means losing focus.
• One owner per metric. Not shared.
• Measurable. Must be a specific number, not subjective.
• Prefer leading indicators. Activity metrics (calls made, demos scheduled) predict results better than lagging metrics (revenue).
• Seat alignment. Cross-reference metric owners against data/accountability.md. Each metric should map to a seat's responsibilities. Flag mismatches, e.g., "This metric seems to fall under [Seat] responsibilities. Should [Seat Owner] own it?". If a metric doesn't fit any seat, it may indicate a gap in the Accountability Chart.

Step 4: Update the File

Add the new metric row to the table in data/scorecard/metrics.md. Show the diff before writing.

Mode: Log Weekly

Use to record this week's numbers.

Step 1: Determine the Week

If the user specified a week, use it. Otherwise, calculate the current ISO week.

Check if data/scorecard/weeks/YYYY-WNN.md already exists. If so, ask: "Week NN already has entries. Update it, or is this a different week?"

Step 2: Read Metric Definitions

Read data/scorecard/metrics.md to get the list of metrics, their goals, and thresholds.

Step 3: Collect Values

For each metric, ask: "What's the value for [Metric Name] this week?"

Accept the value and auto-calculate status:

• Compare against the Green/Red thresholds from the metric definition
• Set status to on_track or off_track

Step 4: Write the Weekly File

Use templates/scorecard-week.md as the template. Write to data/scorecard/weeks/YYYY-WNN.md.

Show the complete entry before writing:

Ask: "Save this week's scorecard?"

Step 5: Flag Off-Track Items

For any off_track metric, note: "Off-track items should be discussed during the L10 meeting. Consider adding to the Issues list if consistently off track."

Mode: Trend Analysis

Use to review metrics over time and spot patterns.

Step 1: Read Weekly Files

Read all files in data/scorecard/weeks/, sorted by week number. Focus on the most recent 13 weeks (one quarter).

Step 2: Build Trend Table

For each metric, show the last 13 weeks:

Weekly Revenue (brad) — Goal: $50K
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
W01: $48K ✗  W02: $51K ✓  W03: $49K ✗  W04: $55K ✓
W05: $52K ✓  W06: $47K ✗  W07: $53K ✓  W08: $50K ✓
W09: $54K ✓  W10: $51K ✓  W11: $48K ✗  W12: $52K ✓
W13: $55K ✓

On track: 9/13 weeks (69%)
Current streak: 1 week on track

Step 3: Flag Concerning Patterns

• 3+ consecutive weeks off-track: "⚠️ [Metric] has been off-track for [N] consecutive weeks. This may be a systemic issue worth escalating."
• Trending down over 4+ weeks: Even if individual weeks are on-track, flag declining trends.
• Consistently off-track (more than 50% of weeks): "Consider whether the goal is realistic, or if there's a deeper issue. Suggest creating an Issue with ceos-ids."

Step 4: Escalation Suggestion

If any metric has been off-track 3+ consecutive weeks:

"Metric '[Name]' has been off-track for [N] consecutive weeks. This pattern suggests a systemic issue rather than a one-time miss. Consider:

1. Creating an Issue to investigate root cause (use ceos-ids)
2. Reviewing whether the goal needs adjustment
3. Discussing in the next L10 meeting"

Output Format

Define Metrics: Updated metrics table with diff. Log Weekly: Completed scorecard table for the week. Trend Analysis: 13-week trend per metric with pattern flags.

Guardrails

• Always show diff before writing. Never modify metrics.md or week files without showing the change and getting approval.
• Don't skip metrics. When logging weekly, prompt for every defined metric. If a value isn't available, record it as missing (don't silently skip).
• Validate week format. Ensure YYYY-WNN is valid (NN between 01-53).
• Respect metric definitions. When logging weekly values, use the thresholds from metrics.md for status calculation. Don't let the user override on_track/off_track manually.
• Escalation, not alarm. Trend flags are suggestions, not commands. Present data objectively and recommend action, but let the user decide.
• Cross-reference during L10. When invoked during an L10 meeting (via ceos-l10), focus on the latest week's status. Save trend analysis for dedicated review sessions.
• Don't auto-invoke other skills. Mention ceos-l10, ceos-ids, and ceos-annual when relevant, but let the user decide when to switch workflows.
• Sensitive data warning. On first use, remind the user: "Scorecard data may contain sensitive business metrics. Use a private repo."

Integration Notes

L10 Meetings (ceos-l10)

• Read: ceos-l10 reads the latest weekly scorecard during Section 3 (Scorecard Review) of the L10 meeting. Each metric owner reports their number and the team discusses any off-track items.
• Suggested flow: Off-track metrics that persist for 3+ weeks should become Issues via ceos-ids.

IDS (ceos-ids)

• Related: Metrics that are consistently off-track may surface as Issues. When escalating, suggest: "This metric has been off-track for [N] weeks. Create an Issue with ceos-ids to investigate the root cause."

Annual Planning (ceos-annual)

• Read: ceos-annual reviews Scorecard metrics during the annual planning session. The team evaluates whether existing metrics are still the right numbers to track and adds or removes metrics for the new year.

Accountability Chart (ceos-accountability)

• Read: ceos-scorecard reads data/accountability.md when defining metrics to validate that metric owners match seat responsibilities. Revenue metrics belong to whoever owns the Sales & Marketing seat; delivery metrics belong to the Delivery seat owner.
• Suggested flow: If a metric owner doesn't map to a seat, suggest: "This metric doesn't align with any seat in the Accountability Chart. Should it belong to the [Seat] owner?"

Read-Only Principle

Other skills read data/scorecard/ for reference. Only ceos-scorecard writes to scorecard files. This preserves a single source of truth for business metrics.