agentleFS
CodeAlive-AICodeAlive-AI/ceo-ai-os/skills/outreach-sender/SKILL.md

outreach-sender

Skill9 starsChanged 6 months agoMITView on GitHub

Send outreach emails and manage follow-up sequences. Bridges cold-email drafting → humanizer polish → CEO Telegram approval → Gmail send. Tracks sends in pipeline.json. For writing email copy, see cold-email. For lifecycle/nurture emails, see email-sequence.

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 8 are distinctive, and 7 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 7

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Before sending, read MEMORY.md for: current wedge, value proposition.

    line 9
  2. Very specific:

    Read memory/pipeline.json for prospect context and outreach history.

    line 10
  3. Very specific:

    Use cold-email skill with prospect context:

    line 32
  4. Very specific:

    Run through humanizer skill to remove AI patterns.

    line 44
  5. Very specific:

    Update prospect in pipeline.json:

    line 80
  6. Very specific:

    If reply at any point → exit sequence, notify CEO, update stage to meeting_booked or nurture

    line 112
  7. Distinctive:

    If bounce → mark email invalid, try alternate contact via find-email-by-name

    line 113
  8. Distinctive:

    If CEO says "skip" → move to nurture

    line 114
  9. Distinctive:

    After breakup email → move to nurture, re-score in 30 days

    line 115
  10. Distinctive:

    When triggered by daily cron:

    line 119
  11. Distinctive:

    When CEO says "send all approved drafts" or when multiple outreach items are ready:

    line 130
  12. Very specific:

    Update pipeline.json for each

    line 134
  13. Distinctive:

    When CEO reports a reply:

    line 139
  14. Distinctive:

    Update prospect stage (meeting_booked, nurture, or closed_lost)

    line 140
  15. Distinctive:

    If meeting intent: trigger meeting-prep skill

    line 142

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

[YOUR_PRODUCT] Context

Before sending, read MEMORY.md for: current wedge, value proposition. Read memory/pipeline.json for prospect context and outreach history. Output in Russian (email content in English unless specified). For chat platforms: show draft for approval, compact format.

Outreach Sender

Orchestrates the full outreach flow: draft → polish → approve → send → track → follow-up.

Outreach Flow (HITL — Human In The Loop)

1. Draft         → cold-email skill writes email based on signal + prospect context
2. Polish        → humanizer skill removes AI patterns
3. Present       → Show draft to CEO in Telegram for approval
4. Approve/Edit  → CEO approves, edits, or rejects
5. Send          → Gmail via Orth OAuth
6. Track         → Record in pipeline.json
7. Schedule      → Set next follow-up date

Step 1: Draft Email

Use cold-email skill with prospect context:

  • Signal that triggered outreach (from pipeline.json)
  • Company context (from company-intel enrichment)
  • Contact info (from lead-enrichment)
  • Choose framework based on signal type:
    • Funding → "Trigger > Insight > Ask"
    • Hiring → "Observation > Problem > Proof > Ask"
    • Pain language → "Question > Value > Ask"
    • Competitor mention → "Story > Bridge > Ask"

Step 2: Polish

Run through humanizer skill to remove AI patterns. Critical for cold email credibility.

Step 3: Present for Approval

Format for Telegram:

Outreach Draft #[N]:

To: [Name] ([Title] @ [Company])
Signal: [type] — [detail]
Score: [FIT+INTENT+TIMING total]

Subject: [subject line]

[email body]

Reply:
 approve — send as-is
 edit: [your changes] — I'll update and re-present
 skip — skip this prospect
 later — defer to tomorrow

Step 4: Send via Gmail

After CEO approval:

orth run gmail /send-email --body '{
  "recipient_email": "contact@company.com",
  "subject": "subject line",
  "body": "email body (plain text)"
}'

Step 5: Track in Pipeline

Update prospect in pipeline.json:

{
  "outreach": [{
    "type": "email",
    "date": "2026-03-28T10:00:00Z",
    "subject": "subject line",
    "status": "sent",
    "template": "trigger-insight-ask",
    "signal_type": "funding"
  }],
  "stage": "outreach_sent",
  "next_action": {
    "type": "follow_up_1",
    "date": "2026-03-31T10:00:00Z",
    "detail": "Follow-up #1 — new angle (case study)"
  }
}

Step 6: Follow-Up Cadence

DayFollow-UpAngleAction
0Initial emailSignal + pain + valueCEO approves
3Follow-up #1Case study or data pointDraft → approve → send
5Follow-up #2Useful resource or insightDraft → approve → send
7Follow-up #3Social proof (similar company)Draft → approve → send
10Breakup emailRespect + door openDraft → approve → send

Rules:

  • Each follow-up adds NEW value — never "just checking in"
  • If reply at any point → exit sequence, notify CEO, update stage to meeting_booked or nurture
  • If bounce → mark email invalid, try alternate contact via find-email-by-name
  • If CEO says "skip" → move to nurture
  • After breakup email → move to nurture, re-score in 30 days

Daily Follow-Up Check (Cron Mode)

When triggered by daily cron:

  1. Read pipeline.json
  2. Find all prospects where next_action.date ≤ today
  3. For each due follow-up: a. Draft next email in sequence (using previous outreach context) b. Present to CEO for approval
  4. Report: "[N] follow-ups due today: [Company1], [Company2], ..."

Batch Mode

When CEO says "send all approved drafts" or when multiple outreach items are ready:

  1. List all approved-but-unsent drafts
  2. Send each via Gmail with 30-second delay between sends
  3. Update pipeline.json for each
  4. Report: "Sent [N] emails. Next follow-ups scheduled."

Reply Handling

When CEO reports a reply:

  1. Update prospect stage (meeting_booked, nurture, or closed_lost)
  2. Cancel remaining follow-ups in sequence
  3. If meeting intent: trigger meeting-prep skill
  4. Log via structured-log

Related Skills

  • cold-email — writes the email drafts
  • humanizer — polishes drafts to remove AI patterns
  • email-sequence — for lifecycle/nurture sequences (not cold outreach)
  • pipeline-manager — stores all outreach tracking data
  • founder-sales — overall sales strategy
  • find-email-by-name — fallback for bounced emails

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.