Sheet 00 / Index
The library.
01 / Builds
Build 01 · Chief of staff · 12 November 2026
How I built an AI chief of staff that drafts my day and never sends a thing
We run a multi-partner consultancy, and none of us has ever had a chief of staff.
- Running since July 2026
- Seven text files and a manual
- Version 1.3.0, MIT licence

Build 02 · AI Marketing Bench · 12 November 2026
How I built a library of 34 marketing skills for AI agents (and proved each one works)
I spent 25 years in marketing, and I didn't trust a single AI skill I downloaded.
- 34 skills across 7 capabilities
- 295 tests, every scorecard published
- MIT licence

02 / Skills
Outbound6 skills
Ideal customer profile, prospect lists, buying committees, prospecting emails and reply handling, with every claim sourced and compliance checked.
Marketing Audit6 skills
A scored Marketing Health report for any website: positioning, acquisition, conversion, brand and measurement.
Brand4 skills
Brand memory that every skill reads, codified voice rules, an anti-AI-writing pass, and a gate that checks drafts against them.
Content Ops6 skills
Briefs, email sequences, LinkedIn content, a pre-publish QA gate, editorial review and content audits, run as a governed workflow.
Paid Ads4 skills
Audit Google Ads and Meta Ads accounts and build new campaigns and ads, with every claim sourced from brand memory.
Team Ops5 skills
Turn proven work into skills, map which work goes to agents, and route intake and approvals.
Claims & Compliance3 skills
Claims substantiation, accessibility review and disclosure checks for regulated teams.
Search1 skill
SEO, GEO and AEO.
Chief of Staff1 skill
A workspace of behaviours for one leader.
03 / The Bench Standard
Built and tested to one standard
Every AI Marketing Bench skill is built and tested to the Bench Standard and ships with its test results. Where a skill hasn’t fully cleared the bar yet, the scorecard says so.
