Claude Skill

pipeline

Triage a lead pipeline and say who needs action today: replied means send the scheduling link, link sent 24h ago without a booking means nudge, awaiting reply means check inbound, going cold means follow up, booked means close out. Ranked and draft-first, so the owner edits befor

LLM Mart · 0 points · 0 views 0 listing impressions 0 install-command copies
Download tonydzi-second-brain-starter-kit-skills_pipeline-7bc2efa.zip · 2 KB

Install

skills CLI npx skills add https://github.com/tonydzi/second-brain-starter-kit/tree/master/skills/pipeline
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install tonydzi-second-brain-starter-kit@llmmart
Git git clone https://github.com/tonydzi/second-brain-starter-kit.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole tonydzi/second-brain-starter-kit collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

/pipeline — work the leads (today's actions)

🧒 When reporting to a non-technical operator: end with a child-simple "In plain words" recap in their language. NEVER inside lead messages. 📖 Operates under the bible skill — outreach codex _Bible-Outreach-MOC. Outbound = send-direct (the operator 2026-06-16 — they edit the sent message after); NO mass auto-blast (pace + personalize per lead). Money / commitments / credentials → escalate.

🖥️ Dashboard first (the operator works visually)

python "$IMPORTS_ROOT/build_pipeline_dashboard.py" → open $OBSIDIAN_VAULT/_Dashboards/Pipeline-Dashboard.html: a kanban by stage (🔥 replied → ⏰ nudge → ⏳ booked → 👀 waiting → ✅ done) + ready drafts (click = copy). View only, it sends nothing. The text walkthrough below is for the actions themselves (send-direct, the operator edits after the fact).

Step 1 — Load live pipeline state (deterministic, ~free)

Read $IMPORTS_ROOT/tg_followups.json (the watcher's live state). Each pending[] lead: lead, chat_id, username, pitch_sent, calendly_sent, replied, booked, booking_confirmed, + a check instruction. Plus calendly_sent_at (for the 24h nudge) and booking_nudge_rule. Deeper history per lead: 04-Projects\crypto\Platinum-CRM\_Platinum-CRM-MOC.md + its lead cards, or /ask --leads "<name>".

Step 2 — Classify each lead → TODAY's action (priority order)

  1. 🔥 Replied, Calendly not sent (replied set, calendly_sent:false) → draft calendly_text to that chat. Warm NOW = top priority.
  2. ⏰ Calendly sent ≥24h, not booked (calendly_sent:true, no booking_confirmed, calendly_sent_at >24h ago) → draft booking_nudge_text (leads forget to book — standing rule).
  3. 👀 Awaiting reply (pitch_sent:true, no replied) → run the lead's check: read recent messages of chat_id (Telegram MCP; see telegram-howto) for a NEW inbound (sender ≠ Tony) after our pitch. If replied → it becomes case 1.
  4. ❄️ Going cold (pitched long ago, no reply, no nudge) → propose ONE soft follow-up, or mark to drop.
  5. ✅ Booked/confirmed → close out; suggest removing from pending. Never re-pitch.

Step 3 — Output the worklist, then act on the operator's go

  • Show a ranked table: lead · state · proposed action · exact draft text.
  • Send each directly (the operator edits the sent message if something's off); NO mass auto-blast — pace + personalize per lead.
  • After sending, UPDATE tg_followups.json (calendly_sent, booking_confirmed, booking_nudge_sent…) so state stays true.
  • Refresh the lead's CRM card (telegram-lead-outreach capture step).
  • End with 🧒 recap.

Guardrails

  • Voice = the operator's words verbatim (their rule); concise, no filler.
  • Money / commitments / credentials → escalate, never autonomous.
  • If tg_followups.json is empty/stale → say so; offer to rebuild from recent Telegram via telegram-lead-outreach ("find + capture").

Like this skill? It is part of second-brain-starter-kit: the second brain we built for ourselves and run every day at Palo Alto AI Research Lab. Install the whole set with npx skills add tonydzi/second-brain-starter-kit. Everything is open source and free, so take what you need.

Flagships worth a look on their own: secondop-panel (a second opinion from a panel of external models), claude-memory-tidy (stop your agent's memory from rotting), telegram-mcp-kit (your own Telegram over MCP in about 15 minutes).

Author: Anton Dziatkovskii, Palo Alto AI Research Lab. Telegram @tonydzi - WhatsApp +1 341 222 9178 - X @Tony_Stef_

Engineers: want to test-drive this setup? Message me. I hand out free starter seeds to engineers who test and report back, and custom skill requests are welcome.

Files (second-brain-starter-kit)
  • SKILL.md 4.5 KB
    ---
    name: pipeline
    description: >-
      Triage a lead pipeline and say who needs action today: replied means send the scheduling link,
      link sent 24h ago without a booking means nudge, awaiting reply means check inbound, going
      cold means follow up, booked means close out. Ranked and draft-first, so the owner edits
      before anything goes out. Triggers: "/pipeline", "work my leads", "who is stuck in the
      funnel".
    license: MIT
    ---
    
    # /pipeline — work the leads (today's actions)
    
    > 🧒 **When reporting to a non-technical operator:** end with a child-simple "In plain words" recap in their language. NEVER inside lead messages.
    > 📖 Operates under the `bible` skill — outreach codex `_Bible-Outreach-MOC`. Outbound = **send-direct (the operator 2026-06-16 — they edit the sent message after)**; NO mass auto-blast (pace + personalize per lead). Money / commitments / credentials → escalate.
    
    ## 🖥️ Dashboard first (the operator works visually)
    `python "$IMPORTS_ROOT/build_pipeline_dashboard.py"` → open `$OBSIDIAN_VAULT/_Dashboards/Pipeline-Dashboard.html`: a kanban by stage (🔥 replied → ⏰ nudge → ⏳ booked → 👀 waiting → ✅ done) + ready drafts (click = copy). View only, it sends nothing. The text walkthrough below is for the actions themselves (send-direct, the operator edits after the fact).
    
    ## Step 1 — Load live pipeline state (deterministic, ~free)
    Read `$IMPORTS_ROOT/tg_followups.json` (the watcher's live state). Each `pending[]` lead: `lead`, `chat_id`, `username`, `pitch_sent`, `calendly_sent`, `replied`, `booked`, `booking_confirmed`, + a `check` instruction. Plus `calendly_sent_at` (for the 24h nudge) and `booking_nudge_rule`.
    Deeper history per lead: `04-Projects\crypto\Platinum-CRM\_Platinum-CRM-MOC.md` + its lead cards, or `/ask --leads "<name>"`.
    
    ## Step 2 — Classify each lead → TODAY's action (priority order)
    1. 🔥 **Replied, Calendly not sent** (`replied` set, `calendly_sent:false`) → draft `calendly_text` to that chat. Warm NOW = top priority.
    2. ⏰ **Calendly sent ≥24h, not booked** (`calendly_sent:true`, no `booking_confirmed`, `calendly_sent_at` >24h ago) → draft `booking_nudge_text` (leads forget to book — standing rule).
    3. 👀 **Awaiting reply** (`pitch_sent:true`, no `replied`) → run the lead's `check`: read recent messages of `chat_id` (Telegram MCP; see `telegram-howto`) for a NEW inbound (sender ≠ Tony) after our pitch. If replied → it becomes case 1.
    4. ❄️ **Going cold** (pitched long ago, no reply, no nudge) → propose ONE soft follow-up, or mark to drop.
    5. ✅ **Booked/confirmed** → close out; suggest removing from `pending`. Never re-pitch.
    
    ## Step 3 — Output the worklist, then act on the operator's go
    - Show a ranked table: **lead · state · proposed action · exact draft text**.
    - Send each directly (the operator edits the sent message if something's off); NO mass auto-blast — pace + personalize per lead.
    - After sending, **UPDATE `tg_followups.json`** (`calendly_sent`, `booking_confirmed`, `booking_nudge_sent`…) so state stays true.
    - Refresh the lead's CRM card (telegram-lead-outreach capture step).
    - End with 🧒 recap.
    
    ## Guardrails
    - Voice = the operator's words **verbatim** (their rule); concise, no filler.
    - Money / commitments / credentials → escalate, never autonomous.
    - If `tg_followups.json` is empty/stale → say so; offer to rebuild from recent Telegram via telegram-lead-outreach ("find + capture").
    
    <!--kit-footer-->
    
    ---
    
    **Like this skill?** It is part of [second-brain-starter-kit](https://github.com/tonydzi/second-brain-starter-kit): the second brain we built for ourselves and run every day at Palo Alto AI Research Lab. Install the whole set with `npx skills add tonydzi/second-brain-starter-kit`. Everything is open source and free, so take what you need.
    
    Flagships worth a look on their own: [secondop-panel](https://github.com/tonydzi/secondop-panel) (a second opinion from a panel of external models), [claude-memory-tidy](https://github.com/tonydzi/claude-memory-tidy) (stop your agent's memory from rotting), [telegram-mcp-kit](https://github.com/tonydzi/telegram-mcp-kit) (your own Telegram over MCP in about 15 minutes).
    
    Author: **Anton Dziatkovskii**, Palo Alto AI Research Lab. Telegram [@tonydzi](https://t.me/tonydzi) - WhatsApp [+1 341 222 9178](https://wa.me/13412229178) - X [@Tony_Stef_](https://x.com/Tony_Stef_)
    
    **Engineers: want to test-drive this setup?** Message me. I hand out free starter seeds to engineers who test and report back, and custom skill requests are welcome.
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.