pipeline-analysislisted

Run a HubSpot RevOps pipeline and conversion analysis over a period (a quarter, a year, a custom or trailing range) and return a diagnosis-first report: verdict (binding constraint A–F with owner and value), reverse-funnel scorecard, New Business full diagnostic (flow plane, diagnosis plane, cohort conversion, velocity, sources), Account Management condensed, reconciliation, method note. Markdown by default; HTML on request. Use when the user says "/pipeline-analysis", "run a pipeline analysis", "revops analysis", "conversion analysis for [period]", or asks how leads, MQLs, and SQLs converted over a period. NOT the monthly board snapshot (that is marketing-monthly) and NOT a single campaign's report (that is campaign-report). Company and CRM ids come from a client context pack chosen per run; nothing client-specific lives in this skill.
risukisu/marketing-skills · ★ 4 · Data & Documents · score 77

Install: claude install-skill risukisu/marketing-skills

# pipeline-analysis — RevOps funnel & conversion analysis A deterministic engine (`compute_funnel.py`, stdlib only) plus a reporting workflow. Read **`definitions.md`** (same folder): it is the source of truth for every metric. `campaign-report` and `marketing-monthly` point at it rather than redefining metrics. **Paths in this file.** `compute_funnel.py`, `definitions.md`, `references/…` are relative to **this skill's folder**. Output paths come from the client pack, never from the current directory. ## Step 0 — Client context (every run, ask before reading) The skill holds no company or CRM data and never uses any without the user's say-so. 1. List the candidate packs: `references/*-context.local.md` in this skill's folder (machine-local, never shipped), plus anything in the launch directory that looks like reporting context (`CLAUDE.md`, `AGENTS.md`, a `context/` folder, files named for pipeline, funnel, targets, or earlier `*-pipeline-conversion-analysis.md` reports). Filenames only; read nothing yet. 2. Ask one question: use this pack / point me at another file / paste the ids / start with none. Wait. 3. Read only what was approved. A pack supplies portal id, pipeline→funnel mapping, stage ids, ticket stages, AM sources, targets, reason map, output root, house style. "None" means Step 1 must collect the ids by hand and the run stops at markdown with generic labels. 4. **No pack for this company yet?** Copy `references/context-pack.TEMPLATE.md` to