Skip to content
Marketing
Skill

/lynqu-outreach

Write and run first-touch outreach to Lynqu leads (sequences, multi-threaded to the committee) sent through your templates on your approval. Requires the Lynqu MCP server.

From plugin
lynqu-ai-toolkit
617 skills

Claude Code

$ npx -y skills add Gravisun/lynqu-ai-toolkit --skill lynqu-outreach --agent claude-code

SKILL.md

lynqu-outreach.SKILL.md
name: lynqu-outreach
description: Write and run first-touch outreach to Lynqu leads (sequences, multi-threaded to the committee) sent through your templates on your approval. Requires the Lynqu MCP server.

Lynqu Cold Outreach

You write the first message to someone who has never spoken to us, and you make the sequence real in Lynqu: the template exists, the sends are logged on the timeline, and every step that hasn't happened yet is a dated task on the lead.

The bar for a first touch is brutal and worth stating: **would this person be glad they opened it?** If the answer is no, no amount of sequence engineering saves it. One specific, verifiable observation about their business beats four paragraphs of value proposition every time.

Invocation

/lynqu outreach <lead id | company | segment | "everything tagged SaaStr-2026">

Step 1: Earn the right to write

Read the lead first. An email that ignores what Lynqu already knows is worse than no email. It proves nobody looked.

  • **`get-lead`**: role, company, source, score, notes, activity. **How they were

captured is the single most important input**: someone who scanned your badge at a booth is not a cold contact, and treating them like one throws away the only warm thing you have

  • The lead's **notes**, returned with `get-lead`: the rep's own words about

the encounter, if there was one

  • **`list-lead-contact-points`**: has anyone already reached out? Two reps

cold-emailing the same person in one week is a real and embarrassing failure

  • **`get-company`** and **`list-opportunities`**: an existing deal on the account

changes this from cold outreach to multi-threading, which is a different message

  • **`search-contacts`**: is this person already known to someone here?
  • **`list-outbox`** with the `lead_id`: has anyone on the team, or an AI

employee, already emailed them, and did they reply? A reply means this is no longer cold

**Stop conditions.** If the lead was contacted in the last 5 days, is already in an active sequence, has replied, or belongs to another rep's open deal, say so and stop. Don't write the email.

Step 2: Qualify before you spend

Outreach is the most expensive thing in this suite: it consumes the one thing you can't get back, which is the prospect's willingness to hear from you again.

If the lead is unscored, run `lynqu-qualify` first. Below the "nurture" band, recommend *not* sending and say why. A rep who sends 20 well-targeted emails beats one who sends 200, and the second one also burns the domain.

Step 3: Pick the angle

One angle per sequence. Not four.

| Angle | Use when | Opening move | |-------|----------|--------------| | **Trigger event** | Hiring, funding, expansion, a launch | Name the event, then the consequence you'd expect | | **Shared context** | Same event, same community, mutual connection | Lead with the shared thing. It's the reason you're not a stranger | | **Observed symptom** | Something public shows the pain | Describe what you noticed, not what you sell | | **Referral** | A named person suggested them | Name them in the first line or it reads as a trick | | **Post-capture** | They scanned your badge or filled a form | Reference the conversation. You have permission. Use it |

The angle must be **verifiable by them**. "I saw you're hiring three field reps" survives scrutiny. "I imagine you're struggling with lead capture" does not, and it tells them you're guessing.

Step 4: Build the sequence

Four to five touches over 2–3 weeks. Every touch must carry something new, a sequence where step 3 is "just bumping this" teaches people to ignore you.

| # | Day | Job | Length | |---|-----|-----|--------| | 1 | 0 | The observation and one specific question | 40 to 70 words | | 2 | +3 | New angle, not a bump: a proof point or a peer example | 30 to 60 | | 3 | +7 | Something useful with no ask: a relevant number, a short teardown | 50 to 80 | | 4 | +12 | Direct: name the ask, make it easy to say no | 25 to 40 | | 5 | +20 | Close the loop. "I'll stop here. Worth a look when {trigger}?" | 20 to 35 |

Rules that hold across all of them: subject lines under 40 characters and never clickbait; one CTA per message; no attachments on a first touch; plain text, no images; and every message must read as though a person wrote it to *this* person.

**Multi-threading.** If `lynqu-contacts` mapped a committee, the champion and the economic buyer get **different messages**: different pain, different length, different ask. Sending the champion's email to the VP is how you lose the champion. Never send identical copy to two people at the same company.

Step 5: Make it real in Lynqu

Show the full sequence, get approval, then:

1. **`list-followup-templates`**: reuse before you create. A template the org already tuned beats a fresh one, and it keeps the brand consistent. When several fit, **`get-followup-performance`** (group `template`) shows which gets replies; trust a ranking only on 20 or more delivered emails, and only while `replies_tracked` is true. AI employees' proposals in the list are drafts, not sendable templates 2. **`create-followup-template`** for a genuinely new, reusable sequence. `scope: team` makes it the org's (needs `follow_ups.manage`, admins by default); `scope: mine` saves it as the user's personal template, used only on their own leads. If team is denied, offer personal, or deliver the copy and say who can save it for the team 3. **`send-followup-now`** for step 1, **only after an explicit yes for this specific send**. Not implied by plan approval, not implied by "looks good" on the draft. The user says send, or nothing sends. It takes a team template or the user's own personal one. It answers "queued", and the server can still cancel it at dispatch (an opt-out, a reply, a booked meeting), so confirm in **`list-outbox`** before reporting it sent 4. **`add-lead-contact-point`** for every remaining step,

Read more
@gravisun
Ships withlynqu-ai-toolkit

Operate your entire lead motion (research → capture → manage → follow up → measure) straight from Claude, ChatGPT, Codex or Cursor. Open-source Claude Skills + MCP for AI lead capture, lead research, sales follow-up, and pipeline management. Powered by Lynqu.

Stars on GitHub

Momentum

stars in total
6
Star history of Gravisun/lynqu-ai-toolkit on GitHub, a chart by star-history.com

Other skills on lynqu-ai-toolkit.

lynqu-contacts

lynqu-contacts

Map the buying committee on an account (economic buyer, champion, blocker) and write it into Lynqu as contacts and participants.…

lynqu-deal-desk

lynqu-deal-desk

Price a Lynqu deal and draft the quote. Browse the price book, put priced lines on a deal, then produce a draft quote for a human…

lynqu-icp

lynqu-icp

Build an Ideal Customer Profile from your own won and lost leads in Lynqu, then encode it as lead scoring rules so every future…

More skills from other plugins.

seo

seo

Comprehensive SEO analysis for any website or business type. Full site audits, single-page analysis, technical SEO (crawlability,…

@agricidaniel@agricidanielInstall Skill
brag

brag

Turn the current project website into a short, polished, shareable launch video using Hyperframes. Use when someone says "/brag",…

brag2w21k
@latent-spaces@latent-spacesInstall Skill
geo-audit

geo-audit

Full website GEO+SEO audit with parallel subagent delegation. Orchestrates a comprehensive Generative Engine Optimization audit…

Questions about lynqu-outreach.

Which coding agents can use lynqu-outreach?
lynqu-outreach installs in Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, OpenCode and Windsurf with the skills CLI, using each agent's flag shown on this page. Flowy's auto-invocation, which fires a skill when your prompt matches, runs in Claude Code only.