Skip to content

Your team's AI starts from zero. Yours doesn't have to.

A shared knowledge base that ChatGPT and Claude can actually read. Built for small technical teams, founders, and indie SaaS: one place for the context your whole team keeps re-explaining.

Free to start, no credit card. Everyone brings their own ChatGPT or Claude.

The shift

Everyone's AI, reading the same notes.

What changes when your team's context lives in one place your AI can read.

Before Hjarni With Hjarni
Every teammate explains context to their own AI Every teammate's AI reads the same team notes
Decisions disappear into Slack Decisions live in a searchable decision log
New hires interrupt senior people for basics New hires ask Claude or ChatGPT first
Runbooks are passive docs Runbooks become instructions the AI can follow

How it works

One shared brain. Every teammate's AI.

No new habits. Each person keeps the AI they already use.

1

Put your team's context in shared folders

Runbooks, decisions, onboarding, customer interviews. Plain Markdown, one folder per kind of context, each with its own AI instructions.

2

Everyone connects their own AI

Each teammate connects ChatGPT or Claude to the team's knowledge base. No bundled AI, no per-seat token bills.

3

Every AI reads and updates the same notes

Ask a question, the AI searches the team's folders and answers from them. Write something back, and the next teammate's AI already knows it.

The Acme Engineering team sidebar in Hjarni, with Team Knowledge Base expanded to show Architecture, Conventions, Customer Interviews, Decisions, Onboarding, Runbooks, and Writing Style as sub-folders.
One shared knowledge base, one folder per kind of context.

See it work

A new hire asks. The team's notes answer.

No senior engineer interrupted. Claude reads the team's Runbooks and Decisions folders and answers from what's written down.

Claude
I'm new here. How do we deploy to production?
hjarni: search (Runbooks)
From the team's Deploy runbook: merge to main, CI runs, then run bin/deploy production. It tags the release and runs migrations before the cutover. Deploys are paused on Fridays.
What if a migration fails halfway?
The runbook says roll back with bin/deploy rollback, then post in the incident channel. The Decisions folder notes why migrations are kept backward compatible for one release.

Claude answers a new hire's deploy question from the team's shared runbook.

What teams keep in Hjarni

The context you keep re-explaining.

Write it down once. Every teammate's AI reads it from then on.

Runbooks the AI can read

Deploy steps, rollback procedures, and incident playbooks. Add a folder instruction and your team's AI becomes the first responder, not a wiki nobody opens at 3am.

Decision logs as institutional memory

Why you picked Postgres, why pricing went flat, what you tried and rejected. When anyone asks "why do we do X this way", the answer comes from the log, not guesswork.

Shared customer interview memory

Thirty interviews in, ask "what did our last ten customers say about onboarding" and get real answers with real quotes, from every teammate's AI.

Onboarding for new hires

The unwritten context lives in three Slack channels and one person's head. Write it once and a new hire's AI reads it on day one.

What makes it different

Simple on purpose, even for teams.

AI-native, not AI-added

The MCP server is built in, not bolted on. Every folder can have its own AI instructions. Claude and ChatGPT follow them.

Personal and shared together

Personal notes stay personal. The team knowledge base is shared. One login, both spaces.

Roles, not all-or-nothing

Give each teammate a viewer, editor, or admin role, from a team-wide default down to individual folders. Sensitive folders stay read-only for most of the team while shared ones stay editable.

No bundled AI

Team members bring their own ChatGPT or Claude. No per-message costs. No token limits per seat. No surprises on the bill.

Roles & permissions

Who can edit what

Everyone on the team can read the shared knowledge base, so every teammate's AI works from the same notes. Roles just decide who can change what.

Each member has a role on every folder: a viewer reads, an editor reads and writes, an admin also manages other members' roles. Set a team-wide default, then override it per folder where you need to. The owner always keeps full control of every folder.

The AI Instructions panel on the Runbooks team folder, showing the Acme Engineering team-wide rule 'Quote sources verbatim. Tag every note with the relevant area.' above the folder rule 'When asked about deploys or incidents, search this folder first and follow the steps exactly. Cite the note title.'
Team-wide rules sit above each folder's own. Every teammate's AI follows both when it reads team notes.

Need SAML SSO, SCIM provisioning, or a per-team audit log? See Hjarni for Teams (Enterprise).

From the people who use it

We run Hjarni on Hjarni. Our roadmap, launch checklists, and customer notes live in one shared knowledge base, and Claude reads and updates it every session instead of me re-explaining where things stand.
Evert Van den Bruel, Founder, Hjarni

Pricing

Per seat. No AI on the bill.

Teams is $20 per seat per month for the shared knowledge base. Seats prorate automatically as people join or leave. Everyone brings the ChatGPT or Claude subscription they already have.

Start free for your first 25 team notes. No card needed.

Read the teams and sharing docs

Common questions

Common questions

Can the same note be both personal and shared?

Each note lives in either your personal space or a team, and you can move it across whenever you need to.

Do all teammates need their own AI subscription?

Everyone uses the AI they already have, like ChatGPT, Claude, or Cursor. Hjarni gives that AI a shared memory rather than bundling one, so each teammate keeps their own subscription.

How does pricing work for teams?

Teams is $20 per seat per month, free for your first 25 team notes. Seats prorate automatically as people join or leave, and the plan covers shared folders, team-level AI instructions, and member management.

Can different teammates set different AI instructions?

Team-level instructions are shared across the team. Personal instructions stay personal.

Can I control who can edit which folders?

Each member has a viewer, editor, or admin role on every folder, resolved from a team-wide default and any per-folder overrides. A viewer reads, an editor also writes, an admin also manages other members' roles. Everyone on the team can read the shared folders; roles decide who can change them. Roles cascade to subfolders, and the team owner always keeps full control.

Which templates fit a team setup?

Developer Project Memory and Customer Interview Repository are the two strongest team starters. Meeting Memory and Writing Style Guide pair well once the team is past the first install.

Start your team

One brain. Every teammate's AI.

Start free with 25 team notes. No credit card required.

Per seat when you grow. Everyone brings their own ChatGPT or Claude.