Everyone's building Claude skills right now.
Most are built for marketers. I wanted to see what one could do for a professional services practice.
So I built a skill around the 15-minute ritual before every client call.
15 minutes per call. 8 calls a week. Two hours piecing together information that already exists.
The skill compresses that into a two-minute brief.
Paste 3-5 recent client emails. Get a seven-section brief that tells you what's moving, what's stalling, and what the client isn't saying out loud.
I've been deploying variations of this inside client firms for months. The pattern is the same everywhere -- information scattered across five tools. The skill turns it into the analysis that takes experience and attention to do well.
It's called the Client Intelligence Brief. I'll send it to you for free.
| Element | v4 | v5 | Why |
|---|---|---|---|
| Tool names in pain list | "Liscio or email" / "Karbon or Financial Cents" | "your CRM or email" / "Slack or Asana" / "your project tracker" | v4 tools (Liscio, Karbon, Financial Cents) are tax/accounting-specific. Primary outbound target is consultants, fractional CFOs, RIAs. Replaced with tools they actually use. "Your CRM" and "your project tracker" stay specific enough to feel real without excluding anyone. |
| Output description | "relationship status, every open item, a sentiment read, and specific next actions" (4 items -- describes v1) | "a seven-section brief that tells you what's moving, what's stalling, and what the client isn't saying out loud" | v2.0 skill has 7 sections. But listing all seven in the post reads like a spec sheet. v5 signals depth ("seven-section") and teases the surprising outputs (velocity, sentiment, risk) through what they REVEAL, not what they're called. The image shows the full section list — that's where the inventory works. |
| P10 line | "Nobody synthesizes it. Now a skill does." | "The skill turns it into the analysis that takes experience and attention to do well." | v4 line was trying to be quotable (P10 flag). Replacement states the actual value: synthesis that requires judgment, not just aggregation. |
| P2 line | "The Client Intelligence Brief. Free. I'll send it to you." | "It's called the Client Intelligence Brief. I'll send it to you for free." | v4 was a three-beat parallel (P2 flag). Rewritten as a natural two-sentence offer. "It's called" provides the name introduction the reader needs. |
| Value proposition | Implied promise: saves you 15 min of scanning. But the skill still requires pasting emails -- scanning is not eliminated. | 15-min ritual is the hook (pain they recognize). Promise shifts to what the brief gives them: workstream grouping, velocity, dual sentiment, risk flags -- analysis that takes experience. | The real value is synthesis quality, not time elimination. The skill doesn't replace the scanning; it replaces the mental work of turning scattered information into actionable intelligence. |
| Hook direction | "Everyone's building Claude skills" → "I wanted to see what happens when you build one for a professional services practice" (3 sentences before the reader sees themselves) | "Everyone's building Claude skills" → merged lines 2-3 into one sentence → "the 15-minute ritual before every client call" arrives by line 3 | Practice Builders ICP is AI-forward, so Claude skills trend hook stays. But compressed the bridge so the reader sees their own pain by sentence 3 instead of sentence 4. |
| Image mockup | 2x2 grid showing 4 outputs (Status, Open Items, Sentiment, Next Actions) | Vertical list showing all 7 sections with gold left borders and section numbers | Image must match v2.0 skill. 7 items don't fit a 2x2 grid. Vertical list with numbered rows is scannable and accurate. |
All 11 patterns checked against final post copy.
| Pattern | Severity | Result | Notes |
|---|---|---|---|
| P1: Twinning | P1 | Pass | No "You don't have X. You have Y" structures. |
| P1: Mirror Reversal | P1 | Pass | No reversed-word constructions. |
| P1: Question-Silence-Revelation | P1 | Pass | No "went quiet" or dramatic pause arcs. |
| P1: Compound Corrections | P1 | Pass | No "It's not just X. It's that..." patterns. |
| P2: Three-Beat Parallels | P2 | Pass | CTA list has 3 items but each has different structure and length (action + parenthetical reason). The pain list has 4 items, not 3, and uses a data-pattern format (timed breakdowns) which qualifies for the exception. v4's "The Client Intelligence Brief. Free. I'll send it to you." was a three-beat -- now rewritten. |
| P2: Not Because X. Because Y. | P2 | Pass | No negation-correction constructions. |
| P2: Over-Validation | P2 | Pass | No "That's exactly..." or evaluative responses. (Post, not DM -- but checked anyway.) |
| P2: Formulaic Setups | P2 | Pass | No "Most people think X. Actually Y." or "Here's the thing" constructions. |
| P2: Trying-to-Be-Quotable (P10) | P2 | Pass | v4's "Nobody synthesizes it. Now a skill does." was flagged -- rewritten to "The skill turns it into the analysis that takes experience and attention to do well." States the value plainly. |
| P3: Dramatic Single-Word Beats | P3 | Pass | No standalone dramatic fragments. |
| P3: Rhetorical Hand-Holding | P3 | Pass | No "Right?" / "Sound familiar?" / "Here's the thing." |
| P3: Identical Openers | P3 | Pass | No three consecutive sentences with the same opening word. Sentence openers vary: "Everyone's" / "Most" / "So" / "15" / "The" / "Paste" / "I've" / "It's". |
8 rules applied to final post copy.
| Rule | Result | Notes |
|---|---|---|
| 1. End Strong | Pass | Key final words: "right now" / "practice" / "client call" / "memory" / "exists" / "brief" / "next actions" / "tools" / "well" / "free." Each sentence ends on the word that carries weight. |
| 2. Define Acronyms | Pass | CRM appears once -- widely understood in the ICP. No other acronyms used. DM appears in CTA context (LinkedIn standard). |
| 3. Economy | Pass | Key cuts from v4: "I wanted to see what happens when you build one for" compressed to "I wanted to see what one could do for" (saves 3 words). "information your tools hold" to "information that already exists" (cleaner, avoids second-person possessive stacking). |
| 4. No Repeats | Pass | No word appears twice within the same sentence. |
| 5. Vary Word Choice | Pass | "skill" appears 3x -- acceptable as the product category name and trend keyword (same finding as v4). "client" appears 3x -- once in the pain list context, once in the output description, once in the product name. Justified: it is the subject of the skill. No other word hits 4+. |
| 6. Precision Vocabulary | Pass | Precision upgrades retained from v4: "compresses" (not "reduces"), "reconstructing" (not "figuring out"), "velocity tracking" (not "progress"), "dual sentiment reads" (not "feelings"). New: "workstream" (not "category"), "prioritized" (not "suggested"). |
| 7. Two-Comma Rule | Pass | The seven-section sentence uses commas in a list -- this is a serial-comma enumeration, not clause stacking. Every non-list sentence has 2 commas max. |
| 8. Kill Adverbs | Pass | Zero adverbs in the post. |
Updated from v4's 2x2 grid to a vertical list showing all 7 sections. Numbered rows with gold left borders. Scannable at feed size.
Why the vertical list works better than v4's 2x2 grid:
Production note: This HTML mockup shows the layout. For the actual 1200x1200 LinkedIn image, recreate in Canva or Figma at exact pixel specs. Gold (#b79d64) left-border on each output row is the visual thread. Dark background (#0a0a0a) stops the scroll in a sea of white LinkedIn cards.
"Everyone's building Claude skills right now. Most are built for marketers. I wanted to see what one could do for a professional services practice."
"Before every client call, you spend 15 minutes piecing together information that already exists. CRM for the last exchange. Slack for updates. Project tracker for task status. Memory for what's still open."
v4 implied the skill saves you 15 minutes of scanning. But the skill requires pasting emails -- it doesn't eliminate the scanning. That's a false promise that breaks trust when someone uses the skill and realizes they still had to go find the emails.
v5 separates the hook from the promise:
| Element | Role | v5 Copy |
|---|---|---|
| Hook | Pain recognition. Gets the reader nodding. | "the 15-minute ritual before every client call" + timed pain list |
| Compression | Time claim that IS accurate. The skill turns emails into a brief in 2 min. That's real. | "The skill compresses that into a two-minute brief." |
| Promise | What the brief actually gives them -- analysis that requires judgment. | "seven sections: orientation, open items grouped by workstream, velocity tracking, dual sentiment reads, risk flags, and prioritized next actions" |
| Authority | Why this skill exists -- Kathryn has seen the pattern. | "produces the analysis that takes experience and attention to do well" |
The reader's takeaway should be: "This skill doesn't just summarize my emails. It gives me the kind of analysis I'd do myself if I had 30 minutes and fresh eyes." That's the real value of the v2.0 brief.