02 · CaptureApex tier
Daily saved-search outreach
Turn a Saved LinkedIn Search into Daily Approved Pipeline with an AI Agent
The flagship capture loop: a saved LinkedIn signal runs every day — search, filter, enrich, draft — then stops and asks you. Only approved outreach sends; everything is logged.
The Loadout, step by step
- See
searchRun the saved signal — the same ICP search, every day. - Decide
enrichEnrich new matches; keep only real ICP fits. - Decide
list_actionsCheck history — never double-touch someone the loop already reached. - Act
send_messageDraft light outreach for the keepers and queue it — send and log only what you approve.
The loadout.md — paste it to your agent
One portable file: config questions first, real MCP tools in the steps, a hard approval gate. Save it as daily-saved-search-outreach.loadout.md or paste it straight into Claude, ChatGPT, or Cursor once Apex is connected.
# Daily saved-search outreach — Loadout ```yaml loadout: daily-saved-search-outreach tier: apex # Apex tier (or an active Apex trial) mcp: https://apex.leadshark.io/mcp ``` **The job:** The flagship capture loop: a saved LinkedIn signal runs every day — search, filter, enrich, draft — then stops and asks you. Only approved outreach sends; everything is logged. ## Before you run — ask the user Only ask what you don't already know — check the conversation, your memory, and the account itself first. Confirm anything you inferred in one line instead of re-interviewing. 1. **What's the saved search / ICP signal to run?** (role, company type, region) 2. **What's the daily cap on new prospects to draft for?** 3. **What's the outreach angle — the offer or reason to start the conversation?** 4. **Who should never be contacted?** (customers, competitors, a suppression list) ## Steps 1. **See** — call `search`: Run the saved signal — the same ICP search, every day. 2. **Decide** — call `enrich`: Enrich new matches; keep only real ICP fits. 3. **Decide** — call `list_actions`: Check history — never double-touch someone the loop already reached. 4. **Act** — call `send_message`: Draft light outreach for the keepers and queue it — send and log only what you approve. ## STOP — approval gate Present everything you found and drafted BEFORE any write. Do not send, schedule, publish, connect, or create automations until the user approves. Edits beat re-drafts: apply their changes, don't start over. ## Never - Write to LinkedIn without explicit approval this run. - Invent facts, numbers, results, or people. - Contact anyone on the do-not-engage list, or double-touch someone (check `list_actions` first). - Retry around a refusal — report what was refused and why. ## Done means - The user approved the drafts (aim for ≤2 edit rounds). - Approved actions are queued or scheduled — verify via `list_actions`. - Report back: what ran, what's staged, and where to monitor it. ## Why this stays account-safe Read-only until the last step. Drafts stage in the approval queue, sends share your daily budgets, and list_actions keeps the loop deduped across days — the same discipline whether it runs once or every morning.
In a hurry? The one-liner
Using Apex, every day: run my saved search for {ICP signal}. Filter out bad fits, enrich the rest, and check list_actions so we never double-touch anyone. Draft light outreach for the top 10 and queue everything for my approval — send and log only what I approve.Run this Loadout with your agent
Connect Claude, ChatGPT, or Cursor to Apex in about five minutes — one-time 24-hour free trial, no credit card.