From file to useful result
How to ask with this skill installed
Installing the skill gives your AI a method. Your request still has to provide the case-specific facts, constraints and expected output.
Describe the decision or deliverable, not just the topic.
Add source material, audience, limits and known facts.
Set format, quality criteria and checks.
Mistakes to avoid
Error 1: Pitching in the connection request The single biggest reason outreach fails. "I help X do Y and would love to show you my services" signals immediately that you're selling. LinkedIn users are trained to ignore anything that sounds like an ad. If your first message mentions your offer before they've accepted, rewrite it.
Error 2: Zero personalization Sending the same template to every prospect is detectable and feels like spam. At minimum, change the person's name, their company, and one specific observation. Generic messages get generic results.
Error 3: No follow-up sequence Most people send one message, get no reply, and move on. Data shows that 25-40% of responses come after the second follow-up. Without a 3-touch sequence, you're leaving most of your pipeline on the table.
Error 4: Ignoring LinkedIn's limits Sending 100 connection requests in a day will get your account restricted. Stay at 50 per week maximum. Spread them across the week — 10 per day is safer than 50 on Monday.
Error 5: Not tracking results If you don't know your acceptance rate, reply rate, and call booking rate, you can't improve. Log every message you send and its outcome in a spreadsheet. Test different hooks monthly.
Error 6: Reaching out without an ICP Sending messages to everyone in your target industry without narrowing by role, seniority, or pain point dilutes your results. "Anyone in tech" is not an ICP. The more specific your ICP, the more relevant your message, and the higher your response rate.
---