The self-maintained context library
for AI-native marketing teams
Everything your agents need to know about your company, kept current and shared across the team.
Reply to Claude…
Every marketing team has the same AI tools, so the output is starting to look the same. What sets you apart is the context and skills behind your agents. Moso keeps it all up to date and available to every agent your team uses.
Capture your marketing context and processes
Bring your context and skills into one structured library, built from the tools you already use.
Keep it up to date
Moso updates it from your tools and market monitoring, and saves every change as a version.
Pleo changed its pricing page
Pleo
market/pleo.md
Saved as version 12
Accessible by every agent
Every agent gets the right context and skills, and you can see what each run used.
ClaudeWrite LinkedIn ads for the November launch
Moso: launch plan, Head of Finance, linkedin-ads
ChatGPTDraft an SEO brief on expense approvals
Moso: buyer questions, claims, seo-brief
CursorBuild the Approvals landing page
Moso: launch plan, brand voice, claims
n8nSend Monday’s market digest
Moso: competitors, news, market-digest
GeminiPrep a battlecard on Pleo
Moso: Pleo, claims, battlecard
A knowledge base
that maintains itself
Moso builds your context from your tools and your market, and keeps it up to date as things change.
Your launch plans, pricing and claims become entries, and stay in sync with the docs they came from.
November launch: Approvals
Launch date: 18 November
Audience: finance leads, 50 to 200 people
Message: approve spend where your team spends
Channels: LinkedIn, email, webinar on 20 Nov
Last edited by Maya, 3 days ago
Synced from Notion
November launch plan
Launch
Fennec Approvals, live 18 November
Audience
Finance leads at companies of 50 to 200 people
Linked to
Head of Finance persona, Claims and proof, Pleo
Used by
linkedin-ads, launch-email, webinar-invite
The right context
for every task
More context makes agents worse. Moso gives them the right context for each task, at the moment they need it.
Agents see a map of your library, then open only the entries a task needs.
Your library, 214 entries
Claude opened 0
Claude saw a map of the library first, then opened only the entries the task needed.
Personas built
from real buyers
Personas built from what buyers say in calls, deals and searches, and updated as they change.
Every persona field links to the calls, deals and searches behind it, with identifiers removed.
Finance lead, fintech, 50 to 200 people04:12
“Honestly, month-end is three days of chasing receipts. If approvals happened when people spend, we’d close in a day.”
Account executive11:40
“What would stop you switching?”
Finance lead11:46
“We’d need to move off our current cards.”
Built from 0 records in Gong, HubSpot and Zendesk
Head of Finance
Names and identifiers removed. Full records stay in Gong, HubSpot and Zendesk.
One skills library
for the whole team
One shared library of your team’s skills, with versions and tests before changes go live.
Every agent on the team uses the same live version of each skill.
Using your team’s battlecard skill. Here’s the draft.
Battlecard: Pleo
Win on a faster close and a full audit trail. Don’t compete on price.
Skills read pricing, claims and personas from the knowledge base when they run.
How teams
use Moso
Moso sits behind your agents, so every piece of work starts from current, shared context and skills.
One source of truth for every agent
Claude, ChatGPT, Cursor and n8n all read the same pricing, claims and positioning.
ReplacesContext pasted into each tool
Change it once, everywhere
Update pricing or positioning in one place, and every agent uses it from its next run.
ReplacesHunting through prompts and files
Skills the team can rely on
Turn one person’s best prompt into a versioned, tested skill the whole team uses.
ReplacesPrompts shared in Slack
Market changes reach your agents
Competitor and category changes land in the library without anyone updating files.
ReplacesCompetitor docs updated by hand
New people and agencies, ready on day one
Give a new hire or agency the full context, with access to only what they need.
ReplacesOnboarding decks and handover calls
Feedback that sticks
A correction from one person improves every agent’s next run, for the whole team.
ReplacesFixing the same mistake every time
Built for the
whole team
Marketing, sales and your agencies work from one library. Nobody installs anything, and you control who can see and edit what.
Access is set per folder in the Fennec workspace.
Nothing to install
Teammates add Moso as a connector in Claude or ChatGPT and sign in. No repo, no files to copy.
Permissions by folder
Give an agency access to the Ads folder only, or let sales read without editing.
Output you can check
Know what every agent used, and keep sensitive records where they are.
Records stay in your tools
Moso keeps short excerpts with names and identifiers removed. Full records stay in Gong, HubSpot or Zendesk.
Every answer is traceable
See which entries and skill version an agent used, and where each fact came from.
Protected entries
Positioning, claims and personas only change when you say so.
Run log
A record of every agent run that used Moso, across tools and teammates.
Bring what you’ve already built
Import your CLAUDE.md, a context repo or your Notion pages. Moso sorts them into entries and skills, and you’re working from them the same day.
Sorted into 9 folders and 6 skills.3 conflicts flagged for review
Pricing
Founding pilot
€99a month per workspace
For the small group of marketing teams in the pilot.
- Set up with you on a call
- Your context library, kept up to date
- Market monitoring of your competitors and category
- Persona studio and skills library
- Works in Claude, ChatGPT, Cursor, Gemini and n8n
- Your whole team in one workspace
Agencies
Let’s talk
A separate workspace for each client, with access for your team and theirs.
Questions
Any agent that supports MCP, including Claude, ChatGPT, Cursor, Gemini and n8n. Your team keeps using the tools they already use.