Run a Health Check on Your LinkedIn Automations and Limits with an AI Agent

# Doctor — Loadout

```yaml
loadout: doctor
tier: pro-plus            # Pro+ or above
mcp: https://apex.leadshark.io/mcp
```

**The job:** Give your LinkedIn a checkup. The Doctor runs a health check on every automation you're running and what actually went out — recent DMs, connects, and comments and their volumes — diagnoses what's drifting or underperforming, and prescribes the fix, all inside safe limits. It never messages anyone; it keeps the engine healthy, not the outreach.

## 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 are you tuning for — more reach, safer pacing, or clearing out stale automations?**
2. **Any automations or limits that are off-limits — don't touch?**
3. **What does 'healthy' look like for you?** (target daily volume, a safety ceiling)

## Steps

1. **See** — call `list_automations`: Pull every automation with its performance — what's running, what's converting, what's drifting.
2. **See** — call `list_actions`: Optional (Apex): review what actually went out — recent DMs, connects, and comments and their volumes — so you're tuning against real numbers, not guesses. Skip on Pro+ and the review still runs.
3. **See** — call `manage_activity_limits`: Optional (Apex): check how much outreach budget is left before you're near a safety ceiling.
4. **Decide** — call `get_automation`: Read the full spec of the laggards — find exactly what to fix: keywords, copy, or status.
5. **Act** — call `edit_automation`: Tune them — refresh keywords and messaging, pause what's stale — each change drafted for your confirmation.
6. **Act** — call `set_daily_dm_limit`: Right-size your daily DM ceiling to stay in the safe zone — you confirm the new number.

## 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

The governance Loadout — and it proves the point by touching zero messages. The Doctor only reads and diagnoses your own system: every automation edit and limit change is drafted for your confirmation — it prescribes, you decide to fill it — nothing is applied silently, and the whole review is logged. Runs on Pro+ (automation + limit management); the Apex reads — recent activity (list_actions) and remaining budget (manage_activity_limits) — add live telemetry and degrade gracefully without them.

Using Apex, run a health check on my LinkedIn system: list every automation with its performance, pull my recent actions to see what actually went out (DMs, connects, comments) and at what volume, and check how much outreach budget I have left before any safety ceiling. Flag the automations underperforming or drifting, read their specs, and draft fixes — updated keywords, refreshed copy, or pausing what's stale. Recommend a right-sized daily DM limit to keep me in the safe zone. Change nothing until I confirm each edit.