Audit

Audit — Score the Cowork AI OS

You score the user's AI business operating system on a 0–100 scale using the 4 C's framework.

Read these first

1. the 4 Cs rubric in the appendix below (25 points per C)
2. about-me/audit-log.md — past entries, so you can compute the delta
3. about-me/about-me.md, business-brain.md, writing-rules.md, memory.md, connections.md, skills-tour.md — to assess the current state
4. about-me/memory-log.md — if it exists, this is the cold archive of older memory entries that rotated out of memory.md. Read it for the full history when assessing Context (C1) — recency of appends, depth of session activity over weeks/months, evidence of pivots. Skip if absent.

What to score

Context (0–25)

• Are all 4 about-me/ files filled in (no ~~placeholder left)?
• Does writing-rules.md have 3+ writing samples?
• Has memory.md been appended to recently (within the last 7 days)?
• Is business-brain.md current (or stale relative to memory entries about pivots/changes)?

Connections (0–25)

• How many of the 7 buckets in connections.md are actually populated (not "none" or empty)?
• How many connectors are enabled in Cowork? (Ask the user if you can't see Settings.)
• Are write actions safely on "Ask before acting"?
• Are there Comms / Tasks / Knowledge connectors, not just Gmail and Calendar?

Capabilities (0–25)

• Are all 7 Cowork AI OS skills installed and functional?
• Are any role plugins from Anthropic installed (productivity, sales, marketing, etc.)?
• Are there any custom skills in skills-tour.md or visible to Claude?
• When was the last custom skill built (per memory.md)?

Cadence (0–25)

• Is the morning brief scheduled and firing?
• Is the weekly tune-up scheduled and firing?
• Are there any other scheduled tasks?
• Is the user's "computer awake" strategy in place (always-on, or accept-when-wake)?

How to compute the score

1. For each C, walk the criteria above.
2. Assign a 0–25 score. Use the rubric in the appendix below to calibrate.
3. Sum to get the total (0–100).
4. Compare to the most recent entry in about-me/audit-log.md to compute deltas.

How to identify the top gap

The "top gap" is not the lowest C. It's the C where the highest-leverage single change lives. Examples:

• If Context is 22/25 (just missing a 3rd writing sample) and Connections is 12/25 (5 buckets unconnected), Connections is the gap — adding ClickUp would unlock daily-brief task awareness, more leverage than a 3rd writing sample.
• If Cadence is 8/25 (no scheduled tasks at all) and everything else is good, Cadence is the gap — getting the morning brief scheduled is huge.

Pick the gap by leverage, not by lowest score.

Output format

## Cowork AI OS Audit — YYYY-MM-DD

**Total: X/100** (delta: ±Y from last audit on YYYY-MM-DD)

| C | Score | Notes |
|---|---|---|
| Context | X/25 | one-sentence summary |
| Connections | X/25 | one-sentence summary |
| Capabilities | X/25 | one-sentence summary |
| Cadence | X/25 | one-sentence summary |

**Top leverage gap:** [Specific thing — "Add a Tasks-bucket connector (ClickUp or Notion) — would unlock daily-brief task pulls"]

**Top 3 things to consider this week:**
1. [Most leverage]
2. [Second]
3. [Third]

Want to act on the top gap right now? Type **fix it**. Want to wait for Friday's tune-up? Type **wait**.

After showing the score

Append a new entry to about-me/audit-log.md using the format in templates/audit-log.md.

If the user says fix it:

• Pretend you're at Gate 3 of the tune-up loop.
• Propose ONE specific change. Get approval. Make the change. Confirm.

If the user says wait or doesn't reply:

• Note in about-me/memory.md that the audit ran and the user deferred action to the next tune-up.
• Then run the auto-archive check per the rules in about-me/memory.md (Auto-archive section). If memory.md now exceeds 30 entries, move the oldest overflow into about-me/memory-log.md. Silent.

Tone

Be specific, never generic. Don't say "your context is decent" — say "your business-brain hasn't been updated since [date], even though memory.md mentions you launched [thing] in March." Cite specifics. Show your work.

---

Appendix: The 4 Cs framework (the audit rubric)

Score the user Cowork AI OS on a 0-100 scale. 25 points per C.

C1 - Context (0-25)

How much does Claude know about the user and their business?

• 0-8: No about-me/ files, or files have ~~placeholders unfilled.
• 9-16: About-me/ has the basics but writing-rules.md has fewer than 3 samples.
• 17-22: All 4 about-me/ files complete. Writing samples real. Memory.md has 5+ entries.
• 23-25: Above plus business-brain.md kept current and memory.md gets appended every session.

C2 - Connections (0-25)

How many of the 7 buckets (Revenue, Customer, Calendar, Comms, Tasks, Meetings, Knowledge) are reachable?

• 0-8: 0-2 buckets reachable.
• 9-16: 3-4 buckets reachable.
• 17-22: 5-6 buckets reachable.
• 23-25: All 7 reachable with safe permissions (writes need approval).

C3 - Capabilities (0-25)

How many things can Claude actually do?

• 0-8: Cowork AI OS only, never run a skill beyond /onboard.
• 9-16: Cowork AI OS skills used regularly. No additional plugins.
• 17-22: Cowork AI OS plus 1-2 Anthropic role plugins. 1-2 custom skills built via /add-skill.
• 23-25: Above plus custom skills get refined every few weeks.

C4 - Cadence (0-25)

How much works on a schedule, vs only when asked?

• 0-8: No scheduled tasks.
• 9-16: Morning brief scheduled. Nothing else.
• 17-22: Morning brief plus weekly tune-up plus at least one custom routine.
• 23-25: Multiple routines firing reliably. Computer-awake reliability solved.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: AutomatedMarketer
• Source: AutomatedMarketer/cowork-ai-os
• License: MIT
• Homepage: https://nunomtavares.com

Install and usage instructions live in the source repository linked above.