Triage Your LinkedIn Inbound with an AI Agent (Signals → First Reply)

# Inbound triage — Loadout

```yaml
loadout: inbound-triage
tier: apex                # Apex tier (or an active Apex trial)
mcp: https://apex.leadshark.io/mcp
```

**The job:** Surface your hottest signals every morning — new commenters, reactors, profile visitors, inbound invites — rank them by fit, and draft the first reply before you've had coffee.

## 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 does a hot inbound lead look like for you?** (titles, company types)
2. **What should the first reply offer — a resource, a call link, just a conversation?**
3. **Which invitations should be auto-accepted vs. left for you?**

## Steps

1. **See** — call `list_signals`: Pull today's hottest inbound leads — who engaged, invited, or messaged.
2. **See** — call `manage_invitations`: List pending connection requests waiting on you.
3. **Decide** — call `enrich`: Enrich the top signals: title, company, ICP fit. Rank who's worth a reply today.
4. **Act** — call `send_message`: Draft a first reply for each keeper and queue it — nothing sends without your approval.

## 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-heavy by design: the only writes are invitation accepts and messages, and messages are staged for approval. Everything shares your daily DM budget and skips the do-not-engage list.

Using Apex every morning, pull my LinkedIn signals and pending invitations. Enrich the new people, rank them by ICP fit (B2B SaaS, founder/GTM titles), and accept invites from fits. For the top 5, draft a personalized first message referencing what they engaged with — put every message in the approval queue, send nothing without my confirmation.