Standalone tool: creates a strategic document in Notion according to the canon
${CLAUDE_PLUGIN_ROOT}/reference/strategy-template.md (9 sections). This is not a PDF for the client,
but a config for all subsequent steps of the content engine (research, weekly plan, production):
sections 4, 6, 7 — machine-readable tables.
Input: $ARGUMENTS — the person's LinkedIn profile + website or company LinkedIn. If something is
missing — ask before starting.
Data collection
- Person's profile (Apify):
apimaestro/linkedin-profile-detail(headline, about, experience) +apimaestro/linkedin-profile-posts(last 50–100 posts). From the posts extract:- the author's real language patterns (for Section 6 — voice profile);
- what has already worked: top posts by engagement (priority for Sections 4 and 7);
- biographical "ammo" facts with numbers (for Section 9). If there are no/few posts — mark it as a "cold account", the voice profile will be [ASSUMPTION] with calibration on the first 10 posts.
- Company: website — via skills
website-scraper+deep-company-analyser(fallback: WebFetch on key pages; SPA fallback: Apifyapify/website-content-crawler). Company-LinkedIn —scrapemint/linkedin-company-employees-scraper(headcount, roles). - Client documents: ask for the ICP doc and value prop. If available — that's the primary source for Sections 2, 5. If not — build assumptions from the scrape and mark each one [ASSUMPTION].
- Reuse of GTM strategy: if the client is in the Notion base "GTM Strategy" (gtm-strategy plugin) — adapt Sections 2, 3, 5 from there, don't duplicate the work.
- Section 5 — offer audit: inventory → LinkedIn-sellability scoring (4 axes × 8 criteria
from
strategy-template.md) → offer ladder gap map → recommendations for new offers [ASSUMPTION]. If there is no offer ladder — first invoke the skilloffer-ladder, then apply the scoring.
Building the document
Fill in all 9 sections per strategy-template.md, with the rules:
- Section 4 (pillars): 3–5 pillars; each tied to a viral lever
(
${CLAUDE_PLUGIN_ROOT}/reference/methodology.md§5) and to proof of competence found in item 1. NO hard quotas — only funnel + day of the week; the balance is maintained by the weekly planner. If the client has little proof — pillars based on backup competence strategies (methodology §2.3). - Rhythm in Section 4: base it on methodology §2.5, but adjust it to the commitment (3 posts/week → Mon/Wed/Fri: TOFU-MOFU-BOFU rotation; 4 → without a second TOFU). Do not fix the formats — they will be determined by testing.
- Section 6 (voice): a list of "what we never write" — at least 5 items, some of them universal (AI clichés), some specific to the person (from the post analysis).
- Example hooks in the pillars — according to the rules
${CLAUDE_PLUGIN_ROOT}/reference/hook-bank.md.
Output (Notion)
- If there is no main "Content Engine" page — create it + DB Idea Pool / Weekly Plans /
Posts per
${CLAUDE_PLUGIN_ROOT}/reference/notion-schema.md. Notify the user of the ID for saving to memory. - Create/update the page «🧭 {Client} · Strategy» with 9 sections.
- At the end of the document — a block "Open Questions": all [ASSUMPTIONS] as a list with a question to the client for each one.
- Seed the Idea Pool: 10–15 starting ideas (2–3 per pillar) with the status
new, Source =matrix-expansion, each with a hook draft and competence proof.
Definition of Done
- All 9 sections filled in; no "TBD" without an [ASSUMPTION] marker.
- Sections 4, 6, 7 — as tables (machine-readable).
- Voice profile contains ≥ 3 signature patterns from real posts (or is marked as a cold account).
- Pillars without quotas: only funnel + day; the rhythm covers the commitment (N posts/week).
- Idea Pool seeded with ≥ 10 ideas, each passes the competence proof gate.
- The "Open Questions" block is formed; a short summary is returned to the user.