Share it with your team

Once your knowledge is in a Brain, the point is that everyone — and everyone’s agents — reads the same copy.

Invite people

From Settings → Members in app.nicia.ai, invite someone by email and pick their role. Accepting the invite creates their organization membership first, and the Brain access they need is granted against that membership — so the review inbox and connecting an agent both work the moment they accept, with nothing to re-authorize.

Connect Claude Code, Codex, Cursor, or any MCP client

export NICIA_API_KEY="…"   # or leave it unset and the CLI prompts for it
npx nicia connect team:acme

Using existing Brain team:acme (name, purpose, and mode left as they are)
Minted a Brain connection.
Claude Code — /Users/priya/.claude.json: added MCP server "nicia-team-acme"

Restart Claude Code and ask it something from team:acme.

There is deliberately no --key flag — a secret on the command line lands in shell history and in ps output for every user on the machine, so the CLI only ever takes it from NICIA_API_KEY or an echo-suppressed prompt.

nicia connect <brain-key> creates the Brain if it does not exist yet, mints a scoped credential for exactly that Brain, and merges one MCP server entry into the target client’s config — every other server already in that file is left alone. --client codex and --client cursor target those clients; claude-code is the default. The credential it mints is named deterministically from the client and this machine’s hostname, so re-running the same command on the same laptop finds and reuses it instead of minting a second one. --org <organizationId> picks the organization when your account belongs to more than one, and --dry-run prints everything it would do — including which file it would touch — without writing anything.

What an agent can do with it

nicia connect mints a scoped credential for exactly this Brain. The tools available track that credential’s capabilities; a default nicia connect grants read and write, which shows up as:

Tool	Does
knowledge_context	Assemble bounded context from accepted knowledge; the response carries a committed read receipt
knowledge_query	Run a structured, snapshot-addressed query against the Brain’s accepted state
knowledge_think	Ask a natural-language question and get a synthesized, cited answer grounded in the Brain’s corpus — restricted to the pages behind accepted claims once the Brain has any; the response carries a committed read receipt naming every source the answer cites
knowledge_submit_evidence	Submit versioned source material — not accepted state on its own, until a mutation cites it
knowledge_submit_mutations	Submit typed knowledge mutations; whether one lands or stages for review follows the Brain’s mode

So “check the handbook before you answer” stops being a thing you paste into a prompt and starts being something the agent can do.

What an agent writes waits for you

A Brain that nicia connect creates is reviewed: a connected agent’s knowledge_submit_mutations call is still typed and canonical, but it stages instead of landing — the mutation waits in the queue carrying the evidence it cites, and nothing changes until somebody accepts it.

Two cases do not get that default. A Brain created over the API with PUT /v1/brains/{handle} is open unless the request states "mode": "reviewed", and connect never changes the mode of a Brain that already exists. It names the mode it found on the line it prints for the Brain, and warns when that mode is open, because a connection minted against an open Brain writes into it unreviewed. Either way the switch is the same call:

curl -X PATCH "$API/v1/brains/team:acme" \
  -H "authorization: Bearer $NICIA_KEY" \
  -H "content-type: application/json" \
  -d '{ "mode": "reviewed" }'

curl "$API/v1/brains/team:acme/changes?status=staged" \
  -H "authorization: Bearer $NICIA_KEY"

Settle it — from the same call, or from the review inbox in app.nicia.ai, which shows the diff and the source side by side:

curl -X POST "$API/v1/brains/team:acme/changes/chg_44a1e0/approve" \
  -H "authorization: Bearer $NICIA_KEY"

mode is the Brain’s own setting, so it governs every connected agent at once, and no credential carries an override of its own. If a pipeline needs to write straight in while agents are reviewed, give it its own Brain. See Keep it good and Review for the full model.

Everyone reads the same thing

The point of the shared copy is that a correction lands once. When Priya fixes the handbook, the next question anyone’s agent asks gets the corrected answer — no re-syncing, no stale local vault, no “which copy is right.”

Next

• Keep it good — review, contributions, and letting agents help maintain it.
• Handing out keys — scoped keys for services and pipelines that don’t run through nicia connect.