podojo-cli 1.6.0

CLI for the Podojo user research platform

podojo-cli

Command-line tool for the Podojo user research platform.

podojo lets UX researchers manage projects, upload and transcribe interviews, build unmoderated user tests, run AI voice interviews, download research videos, cut showreels, and run synthetic test participants — all without leaving the terminal.

Installation

uv tool install podojo-cli

Or with pip:

pip install podojo-cli

Requires Python 3.12+. Some commands need extra tooling:

• showreel requires ffmpeg on your PATH (brew install ffmpeg).
• synth requires the optional synth extra: pip install 'podojo-cli[synth]' then playwright install chromium.

Authentication

Log in once with your Podojo API key — it's stored in your system keychain:

podojo auth login            # prompts for the key
podojo auth login --key XXX  # or pass it directly
podojo auth logout           # remove the stored key

Configuration is read from, in order of precedence:

1. Environment variables — PODOJO_API_KEY, PODOJO_BASE_URL
2. ~/.podojo.toml
3. The system keyring (set by podojo auth login)

Usage

podojo --help            # top-level help
podojo <group> --help    # help for a command group

Projects

podojo projects list
podojo projects create "Checkout Redesign" --brief "Q1 usability study"
podojo projects upload-doc "Checkout Redesign" brief.md --type brief
podojo projects get-doc "Checkout Redesign" --type final -o final-report.md

upload-doc and get-doc accept brief, agent, and final document types. Embedded images are stripped from markdown before upload.

Interviews & transcripts

podojo interviews upload session-03.mp4 --project "Checkout Redesign"
podojo transcripts list "Checkout Redesign"
podojo transcripts download "Checkout Redesign" <batch-id> -o transcript.txt

Videos & showreels

podojo videos list "Checkout Redesign"
podojo videos download <batch-id> -o clip.mp4
podojo showreel create clips.json -o showreel.mp4

The showreel clips.json is a list of clips, each with batch_id, participant, country, topic, start, and end (MM:SS or HH:MM:SS). Each clip gets an auto-generated title card.

Unmoderated user tests

podojo usertests example > my-test.yaml   # print a template to start from
podojo usertests validate my-test.yaml    # check it without creating
podojo usertests create -f my-test.yaml
podojo usertests list
podojo usertests get checkout-usability-v1
podojo usertests update checkout-usability-v1 -f changes.yaml
podojo usertests delete checkout-usability-v1
podojo usertests snippet                  # recorder script for self-hosted prototypes

AI interviews

An AI interview is a self-serve voice conversation: an AI interviewer asks your questions in order and adaptively follows up. One YAML file defines one study:

podojo aiinterviews example > my-interview.yaml   # print a template to start from
podojo aiinterviews validate my-interview.yaml    # check it without creating
podojo aiinterviews create -f my-interview.yaml
podojo aiinterviews list
podojo aiinterviews get checkout-experience-v1
podojo aiinterviews update checkout-experience-v1 -f changes.yaml
podojo aiinterviews delete checkout-experience-v1

create and get print the participant Preview and Live URLs. The frontend base URL can be overridden via ai_interviews_url in ~/.podojo.toml or PODOJO_AI_INTERVIEWS_URL.

A study can open with an optional participant screener: on-screen single-select screening_questions whose options carry qualifies: true flags. Participants must pick a qualifying option on every question; everyone else sees the study's rejection_message and never reaches the voice interview. See podojo aiinterviews example for the exact shape.

Synthetic participants

The synth group drives a Playwright browser through a user test preview so an agent (e.g. Claude Code) can act as a synthetic participant:

podojo synth start checkout-usability-v1   # or a preview URL
podojo synth state
podojo synth click "Get Started"
podojo synth fill "#answer" "It felt slow"
podojo synth advance
podojo synth stop

Links

• Source & issues
• Changelog

License

MIT