agentleFS
link7373link7373/agentic-sales-marketing/.claude/skills/run-outbound/SKILL.md

run-outbound

Skill1 starsChanged 3 months agoMITView on GitHub

Build a targeted prospect list and run a multi-channel outreach sequence — research, personalization, approval, sends, follow-ups, reply triage. The B2B pipeline-filling workflow; run on a standing cadence once strategy approves outbound.

The file

Files
SKILL.md

/run-outbound — Fill the Pipeline

Inputs: knowledge/audience-profiles.md (ICP), knowledge/positioning.md, knowledge/pipeline.md, knowledge/strategy.md (outbound must be an approved channel). Agents: sales-developer leads; copywriter refines messages; tools-operator sends/tracks. Method: knowledge/methods/outbound-prospecting.md; compliance per standards/compliance.md — no purchased lists, honest identification, stop-on-request.

Steps

  1. Build the list. From the ICP: right industry, size, role; prioritize accounts with trigger events (funding, hires, expansion, public complaints about alternatives). Sources: public directories, professional platforms, communities — ethically gathered. Target size per the cadence (quality over volume); exclude opt-outs and existing customers.
  2. Research for relevance. Per prospect (or tight cluster): one specific reason this outreach is relevant to them now. No reason found = off the list.
  3. Design the sequence. templates/outreach-sequence.md: 5–8 touches over 2–4 weeks mixing email and social touches (engage usefully before messaging). Each message: relevance hook → one value point → one low-friction ask. Brief, about their problem, plain sign-off.
  4. Approval gate. Operator approves templates + list once (log the ref). The sequence then runs autonomously, follow-ups included.
  5. Execute on cadence. tools-operator sends per schedule; personalization fields verified before each send batch (a broken merge field costs the reply and the reputation).
  6. Triage replies daily. Positive → respond same day, book the conversation, hand to deal-closer with full context. Objection/brush-off → prepared turnaround (acknowledge → bridge with value → re-ask once), then respect the no. Opt-out → flagged permanently, immediately.
  7. Track ratios. Touches → replies → conversations → qualified opportunities, recorded in knowledge/pipeline.md. Diagnose from ratios: low replies = list or message; low conversions = qualification.
  8. Recycle. Non-responders after the final touch move to a long-cycle nurture list (content, occasional value touches) — never deleted, never spammed.

Before you start

  1. Required inputs: knowledge/strategy.md approved with outbound as a chosen channel; knowledge/audience-profiles.md (ICP) at seeded+; knowledge/positioning.md at seeded+.
  2. Scan knowledge/decision-log.md for tags audience, pipeline, channel; check knowledge/pipeline.md do-not-contact before building any list.
  3. Gate check: the outbound gate (step 4) is absolute — templates + list approved once, then the sequence runs autonomously. Compliance rules 1–5 checked per recipient region.

If things go wrong

  • Sending tool unavailable (Gmail MCP unauthorized, ESP down) → park the approved sequence in knowledge/pending.md with exact authorization steps; never fall back to a method that can't honor opt-outs.
  • Reply-rate collapse or spam complaints → pause sends, diagnose per method §7 (list vs message), fix before resuming; log in the campaign file.
  • Approval doesn't arrive → the list ages; refresh trigger-event research before a late launch rather than sending stale relevance.

Outputs

A running sequence; new pipeline entries; ratio report. Sequence + results live at knowledge/campaigns/YYYY-MM-<slug>.md. Close with the report per templates/skill-report.md.

Writeback

knowledge/pipeline.md; campaign file; approval ref to knowledge/approvals-log.md; learnings (tag: pipeline) to knowledge/decision-log.md.

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.