outbound-engine
Manages the end-to-end strategy and infrastructure setup for cold outbound email campaigns.
Install
mkdir -p .claude/skills/outbound-engine && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/13328" && unzip -o skill.zip -d .claude/skills/outbound-engine && rm skill.zipInstalls to .claude/skills/outbound-engine
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Design, analyze, and optimize cold outbound email campaigns on Instantly. Handles end-to-end ICP definition, expert panel scoring (recursive to 90+), sequence copywriting, infrastructure audit, capacity planning, and an implementation/strategy doc for human review. Use when asked to build cold outbound sequences, optimize cold email, audit an outbound motion, write sales sequences, or design cold email campaigns. Supports "start from scratch" and "optimize existing" modes. Execution skill — it produces copy, math, and a doc; it does not auto-send. Requires Instantly for audit/send features.Key capabilities
- →Define Ideal Customer Profile (ICP) parameters
- →Score email copy using a 10-expert panel recursively
- →Audit Instantly outbound infrastructure for warmup scores
- →Generate capacity planning and weekly metrics targets
- →Draft cold email sequences for human review
- →Produce a strategy document for outbound campaigns
How it works
The skill determines the operating mode (start from scratch or optimize existing), then executes a three-phase workflow: Discovery & Audit, Expert Panel Recursive Scoring, and Deliverables.
Inputs & outputs
When to use outbound-engine
- →Design cold email campaign
- →Audit sending infrastructure
- →Score email sequence copy
About this skill
Outbound Engine
End-to-end cold outbound: ICP definition, expert panel copy scoring (recursive to 90+), infrastructure audit, capacity planning, and an implementation doc. This skill produces a strategy doc and ready-to-load copy for human review — it does not send. See references/workflow.md for the full procedure.
Core Rules
- Determine mode first (existing Instantly account with API key, or starting from scratch) before any other step.
- Run the three-phase workflow: Discovery & Audit → Expert Panel Recursive Scoring → Deliverables.
- Expert panel scoring target is 90/100 — non-negotiable. Iterate until reached.
- Show every scoring round in the final doc — the iteration trail is part of the value.
- Never send or push to the platform autonomously. This skill's output is a doc for human review. Writes/sends happen only after explicit human approval — see the Sending gate below.
- Use capacity math to set realistic volume and pipeline projections.
Operating Modes
Optimize existing — there is a live account to learn from. First questions: (1) Do you have an Instantly API key? (2) Which campaigns are underperforming, and on what metric? (3) Is there historical data to pull? With an API key, run scripts/instantly-audit.py to pull campaigns, account inventory, and warmup scores before touching copy.
Start from scratch — no campaigns yet. First questions: (1) Who is the ICP (titles, industries, company size)? (2) What do you sell, in one jargon-free sentence, and what is the primary offer (audit, trial, demo, call)? (3) What proof points are real and linkable? Use assets/icp-template.md to collect the ICP before writing copy.
The Three Phases (summary)
- Discovery & Audit — Determine mode. If audit mode, run the Instantly audit and flag any account with warmup score <80 or <14 days as NOT ready. Define the ICP, capture business context (offer, real URLs, proof points), and confirm the expert-panel roster. Detail in
references/workflow.md; ICP fields inassets/icp-template.md. - Expert Panel Recursive Scoring — Draft the sequence, then score it with the 10-expert panel (
references/expert-panel.md), each scoring through their own lens. Each round emits a per-panelist score table, an aggregate, the top weaknesses, the edits made, and the revised copy. Score honestly — no padding to 90. Below 90: fix the top 3 weaknesses and run another round. At ≥90: finalize. Copy must followreferences/copy-rules.md(andreferences/instantly-rules.mdfor Instantly variable/deliverability rules). - Deliverables — Produce the strategy doc: brutal-truth analysis, ICP summary, infrastructure status, every scoring round, final copy for every step, implementation plan, capacity math, weekly metric targets, and STOP/START lists. Full doc structure in
references/workflow.md.
What the 90/100 Gate Means
90 is an operational gate, not a vibe. It means the aggregate average across all panelists hit 90 in a single round with no score padded to get there. Below 90, the copy is not done — identify the lowest-scoring lenses, address those specific weaknesses, and re-run. The point of the recursive trail is that each round names a concrete weakness and the edit that fixed it; ship the trail, not just the final copy.
Sending Gate (resolves Rule 5)
scripts/cold-outbound-sender.py and the write paths in scripts/instantly-audit.py exist but are gated. The default deliverable is the strategy doc. Do not invoke the sender or perform any platform write autonomously. Invoke cold-outbound-sender.py only when the human has explicitly confirmed the strategy doc is reviewed and approved AND explicitly asks to send. Always run --dry-run first and show the result before a live send. Never treat "build the campaign" as authorization to send.
Scope Note
The core skill is the three-phase design workflow above. Three scripts ship in this directory but sit outside that workflow and should not run unprompted: scripts/lead-pipeline.py (Apollo→LeadMagic→Instantly sourcing), scripts/competitive-monitor.py (competitor tracking), and scripts/cross-signal-detector.py (multi-source signal detection). Use them only on an explicit, specific request for that function.
Reference Files
| File | Purpose |
|---|---|
references/workflow.md | Full three-phase workflow with scoring criteria and doc structure |
references/capacity-math.md | Capacity formula, weekly metrics targets, add-on recommendations |
references/instantly-rules.md | Instantly variable syntax, sequence structure, deliverability rules |
references/expert-panel.md | Default 10-expert roster with scoring lenses |
references/copy-rules.md | Email copy rules (first sentence, CTA, stats framing) |
assets/icp-template.md | ICP data collection template |
scripts/instantly-audit.py | Pulls campaigns, accounts, warmup scores via Instantly v2 API (write paths gated) |
scripts/cold-outbound-sender.py | Sends approved emails — gated; human approval required (see Sending Gate) |
Related skills
- content-ops — expert-panel quality gate; the recursive 90+ scoring of sequence copy is a content-ops review pass
- marketing-shaper — produces the scoped outbound brief (ICP, goal, structure) this skill executes from
When not to use it
- →When the request is for intake or scoping without a defined ICP or campaign structure
- →When the task involves drafting follow-ups from a specific sales-call context
- →When the request is for A/B test design or statistical analysis of outbound variants
Prerequisites
Limitations
- →It does not auto-send emails or push to the platform autonomously
- →The audit script, send script, and deliverability rules are Instantly-specific
- →It does not handle strategy definition for campaigns
How it compares
This workflow uses a recursive expert panel scoring system to ensure copy quality meets a 90/100 target before generating an implementation document, unlike manual approaches that may lack structured iteration.
Compared to similar skills
outbound-engine side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| outbound-engine (this skill) | 0 | 1mo | Review | Advanced |
| my-email-api | 0 | 1mo | Review | Intermediate |
| brevo-automation | 0 | 1mo | No flags | Beginner |
| himalaya | 50 | 2mo | Review | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
my-email-api
myapihq
>
brevo-automation
aiskillstore
Automate Brevo (Sendinblue) tasks via Rube MCP (Composio): manage email campaigns, create/edit templates, track senders, and monitor campaign performance. Always search tools first for current schemas.
himalaya
openclaw
CLI to manage emails via IMAP/SMTP. Use `himalaya` to list, read, write, reply, forward, search, and organize emails from the terminal. Supports multiple accounts and message composition with MML (MIME Meta Language).
voicemode
mbailey
This skill provides voice interaction capabilities for AI assistants. This skill should be used when users mention voice mode, want to have voice conversations, speak with Claude, check voice service status, or manage voice services like Whisper and Kokoro.
officeemail-management
harperreed
Handle email tasks (checking inbox, drafting replies, managing threads, adding events to calendar). Use when working with emails to prevent common mistakes like broken threading or missing recipients.
campaign-orchestration
ernijsansons
Create and execute personalized multi-channel campaigns across email, LinkedIn, and phone based on intelligence gathering triggers. Use when planning outreach sequences, optimizing engagement strategies, or executing systematic follow-up campaigns.