LinkedIn Agency System
A per-client structure where one client’s proof point cannot end up in another client’s post. The multi-client problem is contamination, and it is a structural problem rather than a discipline problem. This pack ships the structure: what is shared, what is per-client, where the boundary sits, and the approval trail that protects you when a client disputes what was agreed.
No email, no account, no cookie wall. Every file in it is also readable in full on this site.
The folder
linkedin-agency-system/
├── README.md # setup, in order
├── CLAUDE.md
├── AGENTS.md
├── LINKEDIN-LIMITS.md
├── ICP.md
├── BRAND-VOICE.md
├── OUTREACH.md
├── COMPANY.md
├── OBJECTIONS.md
└── PROMPTS/
├── client-onboarding-brief.md
├── multi-client-workflow.md
├── icp-builder.md
├── ghostwriter-instructions.md
└── follow-up-sequence.mdAgency Client Intake, worked
What this pack produces once the files above are filled in — the actual example from this prompt's own page.
Kickoff transcript with a new client selling supply-chain visibility software, 45 minutes, plus 3 LinkedIn posts they had already published.
ICP.md, extract — firmographics as stated on the call: 500-5,000 staff, manufacturing. Trigger: not stated — flagged in the unanswered list rather than guessed.
VOICE.md, extract — from the 3 samples: short sentences, no exclamation marks, one customer number per post. The client described their own voice as "bold and energetic," which disagrees with the samples — the file follows the samples, not the client's description of themselves.
Unanswered — approval turnaround time was never discussed on the call. Flagged as high-damage and sent back to the client the same day, rather than assumed.
About twenty minutes, most of it writing down your ICP
- Download and unzip.
- Fill the _shared files once, for the agency.
- Run the onboarding prompt per client to generate their folder.
- Create one Project per client and attach _shared plus that client only.
What you do with it once it is set up
- Onboard — Kickoff transcript in, four context files and a gap list out.
- Isolate — One Project per client, shared files shared, client files never reused.
- Plan — Capacity computed per sending account, not per client.
- Produce — Research, draft, review — three roles with defined handoffs.
- Approve — A traceable trail with a deadline and a default.
Every file, in full, on this site
Nothing in the pack is hidden behind the download. If you would rather take three files than fifteen, take three.
- Context file
CLAUDE.md
An agent that reads the right files in the right order and knows what it is not allowed to do.
- Context file
AGENTS.md
Three narrow agents with defined handoffs instead of one conversation that drifts.
- Context file
LINKEDIN-LIMITS.md
Your assistant stops proposing 300 connection requests a week, because it now knows the ceiling is about 100.
- Context file
ICP.md
Every prompt you run afterwards knows who you sell to, in terms LinkedIn can filter on.
- Context file
BRAND-VOICE.md
Drafts arrive in your voice on the first attempt instead of after four rounds of "less corporate".
- Context file
OUTREACH.md
Every message an assistant writes comes out at the right length, in the right shape, with the openers you hate already banned.
- Context file
COMPANY.md
The assistant stops describing your company the way its training data describes companies in your category.
- Context file
OBJECTIONS.md
The assistant answers the objection you actually get, in the way that has worked, instead of improvising.
- Prompt
Agency Client Intake
A filled ICP, voice and product brief per client, plus an explicit list of what the kickoff failed to establish.
- Prompt
Multi-Client Workflow
A per-client folder and Project structure, a weekly routine, and the isolation rules that keep the work clean.
- Prompt
ICP Builder
An ICP written in the vocabulary LinkedIn search understands, derived from deals you actually closed rather than from who you wish would buy.
- Prompt
LinkedIn Ghostwriter — Project Instruction
A standing instruction you paste once into a Claude Project or ChatGPT Project, after which every post you ask for arrives in your voice and with a critique already applied.
- Prompt
Follow-up Sequence Builder
A three-to-four touch sequence with dates, channels, a stop rule, and a capacity check against what your plan can actually send.
Use it for your work and your clients' work
Use every prompt, file and pack for your own company or for clients you are paid by, commercially, modified as much as you like. Put them in your Claude Project, your ChatGPT Project, your Cursor workspace or your agent repo. Share a filled-in copy with your team.
Keep the attribution line at the bottom of any file you redistribute unchanged, and credit Linkediz with a link wherever you republish a resource in full.
Repackage the library or a pack as your own product, paid or free, or scrape the section to rebuild it elsewhere. The downloads exist so you do not have to.
These are writing and research aids. They do not know your customers, they can be wrong, and an assistant will state a fabricated fact as confidently as a true one. Read what comes out before you send it to a person.
Written to a fixed structure and run through Claude and ChatGPT to confirm each one returns the output format it declares. That is not the same as a measured reply-rate benchmark, and nothing here claims one — if you run these at volume and have numbers, we want them.
The reference behind this pack
Per-seat pricing, dated, for the engagement maths.
The constraint most agency capacity plans miss.