# Operations Meeting Prep > Researches a meeting attendee, builds a personalized 12-step sales script, and generates a value-driven pitch deck as PDF. Use this skill when the user says "prep for my meeting", "research [name]", "meeting prep", "get ready for my call with [name]", or provides a name after the morning routine asks about meetings. Triggered as optional last step in /morning after the user confirms a meeting. - Skill: `sabahudin-web/operations-meeting-prep` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add sabahudin-web/operations-meeting-prep` - Raw SKILL.md: https://api.skillmd.com/api/skills/sabahudin-web/operations-meeting-prep/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Docs & Writing - Author: sabahudin-web (https://skillmd.com/u/sabahudin-web) - Updated: 2026-09-22 - Page: https://skillmd.com/skills/sabahudin-web/operations-meeting-prep --- # operations-meeting-prep One job: walk into every sales call fully prepared. That means deep LinkedIn enrichment + 30-day web research on the person, a personalized 12-step sales script, and a value-driven pitch deck tailored to that specific person — all stored in Notion under "Meeting Prep Report & Sales Script". **Automation level:** AUTONOMOUS — runs end-to-end without pausing. When inputs come from a calendar booking (via `/morning`), the LinkedIn URL and meeting time are auto-extracted. Offer fit is auto-picked using the scoring rubric below, with ambiguity flagged in the output rather than gated on user confirmation. **Routine position:** /morning step 7. Auto-runs once per Discovery Call booked for today. Also runs standalone when called directly with a name + LinkedIn URL. --- ## Setup Required Replace these placeholders before running: | Placeholder | What to replace it with | |-------------|------------------------| | `YOUR_MEETING_PREP_NOTION_DB_ID` | Your "Meeting Prep" Notion database ID | | `[YOUR_PROJECT_PATH]` | Your local project root path (e.g. `/Users/yourname/Documents/my-claude-os`) | --- ## Context to read before running 1. Read `context/offers.md` — know all your offers, their ideal clients, pricing, and positioning angles. 2. Read `context/icp.md` — know the ICP profile so you can spot fit or misfit quickly. 3. Read `references/sales-script-template.md` — the 12-step framework you'll personalize. --- ## Steps ### Part 1 — Get the person's name, LinkedIn URL, and conversation history **Routine mode (called by /morning step 7):** Inputs are pre-extracted from a calendar booking. The orchestrator passes: - `name` (attendee full name) - `linkedin_url` (from the booking form's LinkedIn URL field) - `scheduled_time` (today's call time) - `dm_conversation` (full DM history with this lead, if available via your LinkedIn outreach tool) If `dm_conversation` is non-empty, treat it as primary context — the lead's own words from DMs are the strongest signal for "why did they book the call". **Standalone mode (user calls skill directly):** 1. Take the name (and LinkedIn URL if provided) from the user. - If no name: ask "Who are you meeting with? Give me their name — and their LinkedIn URL if you have it." - If LinkedIn URL is provided: skip Part 2, go straight to Part 3. ### Part 2 — Find their LinkedIn profile (only if no URL given) 2. Try Playwright first: - Tool: `mcp__playwright__browser_navigate` - URL: `https://www.linkedin.com/search/results/people/?keywords=[FirstName LastName]` - Tool: `mcp__playwright__browser_snapshot` - Extract the most likely profile URL by name match. 3. If Playwright fails: use Relay as fallback: - Tool: `mcp__claude_ai_Relay_APP_Claude_Code__search_linkedin_for_profile` - firstName, lastName, company (if known) - Extract the profile URL from the result. 4. If no profile found: note "LinkedIn profile not found — enrichment limited" and continue with name-only context. ### Part 3 — Enrich the person (LinkedIn enrichment tool) 5. Get their full LinkedIn profile using your LinkedIn enrichment tool (e.g. Trigify CLI: `trigify profile enrich --profile-url `). - Extract: current role, company, headline, about/summary, experience, education, certifications. - If this returns empty or errors: note the failure and continue with whatever partial data is available. 6. Get their company profile: - Extract from the enrichment output above. If company fields are missing: note "Company profile unavailable" and continue. 7. Get their recent LinkedIn posts: - Use your enrichment tool (e.g. `trigify profile posts --profile-url --page 0`) - Get last 3–5 posts. Extract: topics, themes, language used, pain points mentioned. - If this returns empty: note "No posts available" and continue. ### Part 4 — 30-day research (web search) 8. Run targeted web searches to find anything notable about this person or their company in the last 30 days: - Tool: `WebSearch` - Searches to run: - `"[Full Name]" podcast OR interview OR keynote 2026` - `"[Company Name]" funding OR acquisition OR launch OR news 2026` - `"[Full Name]" site:linkedin.com` - `"[Company Name]" [industry keyword] 2026` - Extract: podcast appearances, published articles, funding rounds, product launches, company news. - If nothing found in last 30 days: note "No recent news found" and skip this section in the brief. ### Part 5 — Build the research brief 9. Read `context/offers.md` and identify which 1–2 offers fit this person best based on their role, company size, industry, and pain signals. 10. Write the **Research Brief** (concise — short bullets, no paragraphs): - **Who they are:** Role at Company — 1–2 lines max - **Company:** What it does, size, industry — 1 line - **Right now:** What they're focused on (from posts/activity/news) — 1–2 bullets max, skip if no data - **Recent signal:** Anything notable from the last 30 days — 1 bullet, skip if nothing found - **Pain signal:** Visible problem or frustration based on profile + company context — 1 line - **Best offer fit:** Offer name + one-sentence reasoning - **Call type flag:** e.g., "Employee, not a founder — confirm budget authority early" or "Warm inbound — likely already sold" ### Part 6 — Build the personalized sales script 11. Run a silent **profile analysis** before writing. Think through: - Most relevant aspects of their background (industry, title, experience, seniority) - Specific signals from LinkedIn + research that can inform the script - **Conversation analysis (if present):** What resonated in DMs? What problem did they articulate? Why did they accept the call? - Likely pain points based on their role and industry - Which offer maps most naturally and why - Likely objections (price, timing, DIY) 12. Generate the **Personalized Sales Script** following the 12-step framework from `references/sales-script-template.md`: - **Step 1 (Preparation):** 3–5 specific prep items drawn from their profile and research. - **Step 2 (Micro Rapport):** 2 rapport questions drawn from their background or career path. Never reference LinkedIn directly — use the insight, not the source. - **Step 3 (Agenda & Alpha Frame):** Copy from template — no personalization needed. - **Step 4 (Primary Motive):** 2–3 pain-discovery questions tailored to their specific role and situation. - **Steps 5–6 (Discovery + Pain):** Adapt questions to their business type, team structure, acquisition/growth context. - **Steps 7–9 (Goal, Defeat, Conviction):** Copy from template — no personalization needed. - **Steps 10–11 (Pitch):** Name the offer to lead with. Write a 1-sentence positioning line using their own language and context. - **Step 12 (Pricing):** State full price, offer same-day decision bonus, stay silent. If they say "too expensive": *"Sorry that my price exceeds your value."* — then stay silent. ### Part 7 — Generate the pitch deck 13. Auto-pick the best-fit offer using a scoring rubric. Score every offer in `context/offers.md` against the lead on: - **Role fit** (0–3): how well the prospect's title matches the offer's ICP role - **Company fit** (0–3): company stage/size fit relative to offer's target - **Pain signal match** (0–3): explicit pain in their posts/profile/recent news that the offer solves - **Budget fit** (0–2): seniority + company stage suggest they can afford this offer - **Recent intent signal** (0–2): bookings, podcasts, posts hinting at the offer's space - Total possible = 13 per offer. Highest score wins. **Ambiguity rule:** if the top two offers are within 1 point of each other, pick the highest-scoring one and add a banner: > ⚠ Offer fit ambiguous between **[Top Offer]** and **[Second Offer]** (scores X vs Y). Glance before the call and override the deck if the second one feels closer. 14. Run `offer-build-pitch-deck` with these inputs: - Person: full name, role, company - Enriched profile summary (from Part 3) - Research brief (from Part 5) - Sales script (from Part 6) - Confirmed offer (name, pricing, deliverables from `context/offers.md`) - Deck requirements: - 8–15 pages max - Value and outcome driven — sell what they GET and HOW they achieve it with you, not features - Personalized with their name and company context throughout - Structure: Problem they have → Cost of not fixing it → The outcome they want → How we get there together → Offer details → Pricing → Next step - Save pitch deck to: `inbox/outputs/pdf/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].pdf` > **Note:** `offer-build-pitch-deck` is included in this starter pack under `skills/offer-build-pitch-deck/`. Install and configure it first. 15. **Pitch Deck files:** After the PDF is generated, note both outputs: - PDF path: `inbox/outputs/pdf/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].pdf` - HTML URL: `file:///[YOUR_PROJECT_PATH]/inbox/outputs/html/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].html` ### Part 8 — Save and push to Notion 16. Save the full output to: `inbox/outputs/md/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].md` Contents: - Research brief - Full personalized sales script - Reference to pitch deck PDF path 17. Push to the **Meeting Prep Report & Sales Script** database: - Tool: `mcp__claude_ai_Notion__notion-create-pages` - parent: `{"type": "data_source_id", "data_source_id": "YOUR_MEETING_PREP_NOTION_DB_ID"}` - Properties: - `Day & Date`: `YYYY-MM-DD — [First Last] call` - `date:Date:start`: today's date in YYYY-MM-DD format - `Research Report`: the concise research brief text - `Pitch Deck`: `file:///[YOUR_PROJECT_PATH]/inbox/outputs/html/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].html` - Page body content: the full personalized 12-step sales script 18. Print one-line summary: "Meeting prep ready for [Name]. Research brief + sales script + pitch deck saved to inbox and Notion." --- ## Output Format ```markdown # Meeting Prep — [First Last] — YYYY-MM-DD > ⚠️ [Call type flag — e.g., "Employee, not a founder — confirm budget authority early"] --- ## Research Brief - **Who they are:** [Role at Company — 1–2 lines] - **Company:** [What it does, size, industry] - **Right now:** [What they're focused on — skip if no data] - **Recent signal:** [Podcast/funding/news from last 30 days — skip if none] - **Pain signal:** [Visible problem or gap] - **Best offer fit:** [Offer name] — [1-sentence why] --- ## Personalized Sales Script ### Step 1: Preparation ... ### Step 12: Pricing ... --- ## Pitch Deck PDF: `inbox/outputs/pdf/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].pdf` HTML (interactive): [Open deck](file:///[YOUR_PROJECT_PATH]/inbox/outputs/html/YYYY-MM-DD-operations-meeting-prep-[firstname-lastname].html) ``` --- ## Edge Cases - **LinkedIn URL provided directly:** Skip Part 2 — go straight to enrichment. - **Enrichment returns empty profile:** Use whatever partial data is available. Still generate the full script with role/company-informed questions. Note "Enrichment partial" at top. - **No recent posts:** Skip "right now" bullet in brief. - **No recent news found (WebSearch):** Skip "recent signal" bullet. - **Offer fit ambiguous (top two within 1 point):** Pick the highest-scoring offer, generate the deck for it, AND add the ambiguity banner. Never pause. - **Internal meeting (team call, known client check-in):** Output "Internal meeting — no prep needed." Skip all enrichment. - **offer-build-pitch-deck skill not available:** Note the failure, save the research brief + script only. - **Notion push fails:** Save to `inbox/outputs/` only. Tell user: "Notion push failed — output saved locally."