# Outreach > Outreach Emails - Skill: `matteotitta/outreach` (Agent Skill) - Install (CLI): `npx skillmds@latest add matteotitta/outreach` - Raw SKILL.md: https://api.skillmd.com/api/skills/matteotitta/outreach/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Productivity - Author: matteotitta (https://skillmd.com/u/matteotitta) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/matteotitta/outreach --- # Outreach Emails Generate personalized outreach and follow-up emails for Genesys Growth that convert prospects into clients. Every email demonstrates operator expertise, references specific prospect context, and leads with value. Three modes: single-email (default), generator (build campaign prompt), runner (batch-generate from CSV). ## Doctrine inherited (Step 7 — 0626 rollout) Output complies with: - [`output-tenets.md`](../../../../../../rules/output-tenets.md) — the seven tenets - [`output-simplicity.md`](../../../../../../rules/output-simplicity.md) — length caps, three-layer source placement, robot-tells ban - [`outbound-research-hygiene.md`](../../../../../../rules/outbound-research-hygiene.md) — dated signals, no stale references, no prior-job hooks - Step 6 calibration: see [[feedback_execution_doctrine_refinements_step6]] **Refinements applied to this skill:** | Code | Refinement | How it lands in outreach-emails | |---|---|---| | **R1** | Source placement (three layers) | Emails are **end-customer-facing**. **No sources block.** No `[VERIFIED:...]` tags in body, no footer "Sources:" list. Research citations live in the working draft for QA only; stripped before send. | | **R3** | Product-update tone | When pitching a capability or update, frame as "we shipped X to address Y" — not "we are thrilled to announce." Applies to single-email and runner modes. | | **R6** | CTA hierarchy | Cold/market-facing → sign-up or discovery-call primary. Blog as fallback for prospects not ready. Never both as primary. Warm follow-up to existing pipeline → product-action CTA. | | **R9** | Action-oriented section names | Production-doc sections (Subject / Opener / Value / Proof / CTA / Sign-off) already verb-led. Preserve. Don't rename to status-oriented variants. | ## When to run Trigger on: "write outreach email", "cold email", "follow-up email", "post-call follow-up", "outreach sequence", "warm intro", "referral email", "re-engagement email", "build a campaign prompt", "batch-generate emails from CSV". For campaign mode (5+ prospects against same ICP), see the premium reference. Do NOT use for marketing/nurture emails (use `lifecycle-marketing`), LinkedIn content (use `linkedin-content`), client deliverable emails, or internal comms. ## Inputs Required (at least one): prospect name/company, website URL, conversation notes, LinkedIn profile, or discovery call summary. Optional: referrer context, specific service interest, timeline/urgency. Auto-loaded: client CLAUDE.md voice section. Auto-fetched: Gmail threads, Drive proposals, Calendar history. Full input table + validation checklist in the premium reference. **Recommended for cold mode (load-bearing for the connector-opener doctrine):** - **Peer customer(s):** 1–2 named logos that Genesys (or the client) has worked with that are structurally similar to the prospect's company. Without this, the connector-opener falls back to generic "we work with similar firms" phrasing that tanks reply rates. - **Wedge:** one sentence on why the engagement won — the workflow or capability that maps from peer customer to prospect. If either is missing for cold mode, ask before drafting. Don't hand-wave. For campaign-scale outreach, run `/deepline-enrich` first to validate emails (waterfall across 15+ providers). Single-prospect emails: Apollo MCP enrichment is sufficient. ## Steps 1. **Mode select.** Default single-email; switch to generator for campaigns (5+ prospects, same ICP); switch to runner when applying an existing prompt to a CSV. See the premium reference. 2. **Validate inputs** per the premium reference Input section. If missing, ask for prospect name/company; offer Gmail search. 3. **Phase 1 — Context gathering.** Gmail/Calendar/Drive search → relationship timeline → email type (cold/warm/post-discovery/proposal/re-engagement). Company research: site, pricing, about, funding, LinkedIn. Detail in the premium reference. 4. **Phase 2 — GTM gap analysis.** Assess against Genesys ICP fit (design test, PMM test, persona pages, launch test, founder LinkedIn). Find personalization hooks (funding, hiring, posts, launches, competitor moves). Apply recency rule: 0–4 weeks primary, 4–8 cautious, 8+ skip. 5. **Phase 3 — Drafting.** Select structure: **for cold-with-peer-customer apply the premium reference (6 structural beats + A/B Loom-vs-POC arm + locked voice rules)**; for post-discovery / re-engagement use templates in the premium reference. Lead with specific observation. First-person singular. Operator voice. No forbidden phrases (the premium reference). Apply `.claude/rules/outbound-research-hygiene.md` to every research-derived signal in the draft (no >12mo, no prior-job hooks, dated not "recently", current-company-state only, sourced numbers only). Verify word counts: cold 150 / post-discovery 250 / follow-up 100 / re-engagement 100. Generate 2–3 subject variants. 6. **Self-evaluate.** Specific opening? "I" not "we"? No forbidden phrases? Hook ≤4 weeks old? Mark gaps `[Need to verify: X]`. 7. **Quality gate** per the premium reference (content, evidence, format checklists + anti-hallucination guardrails). 8. **Format output** per the premium reference (relationship context, sources, subject options, email body, word count, personalization notes, "if they reply" thread, iteration prompts). 9. **Suggest chain.** Follow-up sequence (no response in N days), `/proposal` (positive response), `/company-context` (deeper research). 10. **If campaign mode:** generator outputs a prompt artifact to `{client}/sales/campaigns/{campaign-name}/`; runner takes prompt + CSV, validates per-row variables, applies quality gates per row, outputs batch with skip log. Full spec in the premium reference. ## What good looks like ### Evaluations - Opens with specific observation (not generic intro) - Under word limit for type (150 / 250 / 100 / 100) - Single, clear, low-friction CTA - First-person singular ("I" not "we") throughout - No forbidden phrases ("hope this finds you", "touch base", "synergy", "circle back") - Hook recent (≤4 weeks) — skip if stale - Subject line specific, ≤50 characters - All claims traceable to research or conversation; gaps marked `[Need to verify: X]` - Relationship context accurate; personalization notes explain specificity ## Final ship gate Run `/premortem --output` before ship. See [`/premortem` skill](../../../../../meta/orchestration/premortem/SKILL.md) for the 5 execution domains (will-it-resonate / will-it-convert / will-it-stay-on-brand / will-stakeholder-push-back / will-it-degrade-over-time) and output template. Trivial-case escape: `## Premortem\nNo failure modes — trivial change` satisfies the contract for genuinely trivial outputs. --- ## Persuasion & stickiness pass Output complies with [persuasion-and-stickiness.md](../../../../../../rules/persuasion-and-stickiness.md) — Cialdini's 7 persuasion levers + Heath's SUCCESs. Deploy the 1-2 Cialdini levers that fit the reader's barrier (never all seven; every lever must be TRUE), run the SUCCESs diagnostic (Simple / Unexpected / Concrete / Credible / Emotional / Stories) over the near-final draft, then the rule's pre-ship gate.