cliq
How it works

A team you can put on any repeatable workflow.

You bring the ask. cliq brings the specialists. Here's what happens between the two — and why the output is consistent, reviewable, and ready to ship.

The flow

One brief in. A finished deliverable out.

Each team is a coordinated group of AI specialists. Each specialist does one job well — then hands off to the next.

briefYour askplannerscopes the workresearchergathers contextmakerproduces the draftreview gatePASS / ROUTE / ESCALATEdeliveryship · publish · notify
01

You give a brief

An ask, a ticket, a topic — whatever you'd hand to a colleague.

02

Specialists go to work

Each agent has one job and the context it needs to do it well.

03

Reviews happen along the way

Output is gated. If it's not good enough, the team keeps going.

04

You get the finished work

Delivered where you want it — Slack, a Drive folder, a Jira ticket, a GitHub PR.

Specialists, not generalists

One agent trying everything is why AI output feels uneven.

cliq teams are built from focused roles — the kind of specialization you'd expect from a real team. Each role has one job, one set of inputs, one deliverable.

🧭

Planners

Scope the work. Set success criteria. Sequence the steps.

🔎

Researchers

Gather context from sources, CRMs, databases, and past work.

🧪

Analysts

Make sense of what was gathered. Extract what matters.

✍️

Makers

Produce the actual deliverable — draft, reply, proposal, code.

🧑‍⚖️

Reviewers

Approve, send back for revision, or escalate to a human.

📤

Deliverers

Ship the result to the right place — inbox, channel, system of record.

Designed for trust

Every run is reviewable. Every output is explainable.

AI teams are only useful if you can trust what they produce. cliq is built so you always know what happened and why.

Quality gates

Work doesn't move forward until a reviewer signs off. If quality isn't there, the team iterates — automatically.

Human approval where it matters

Set the bar. cliq escalates edge cases to a human with a one-click approve, redirect, or reject.

Full audit trail

See every step, every input, every output. Replay any run. Explain any decision. Nothing is a black box.

Meets your stack

Reads from where your work lives. Delivers where your team is.

cliq is not a new place to check. It's a team that shows up in the systems you already use.

Reads context from

The places your knowledge and data already sit — so the team starts with everything it needs.

Google DriveMicrosoft 365Jira ticketsAny URLAny HTTP APIYour own scripts

Delivers results to

Where your team actually works — no new inbox, no new tab.

Slack channelsJira ticketsGoogle DriveMicrosoft 365GitHub PRsBitbucket PRs
Under the hood · for the curious

Built on primitives you can inspect.

Most users never need to see this. But for platform teams, power users, and skeptics: here's exactly how a cliq team is put together.

Roles

Each specialist is a focused instruction set with explicit inputs and deliverables. Versioned, reusable, reviewable.

Pipeline

Stages run in sequence or in parallel with declared dependencies. cliq handles scheduling, retries, and recovery.

Hand-offs

Every output is a durable artifact. Nothing lives in hidden state. You can diff, share, or replay any run.

Gates

Decision points that produce a verdict: continue, iterate, or escalate. Safety caps prevent runaway loops.

example team
name: weekly-briefworkflow:  - planner  - researcher (depends on planner)  - analyst    (depends on researcher)  - writer     (depends on analyst)  - reviewer   (gate: PASS / ROUTE / ESCALATE)  - delivery   (post to Slack)

A real team definition is a short configuration plus a markdown file per role. Teams live in your project, version with your code, and can be shared on CliqHub.

Ready to ship with a team?

Ready to put a team on it?

Install cliq, pick a ready-made team, and see how fast consistent, reviewable work can move.