The h3 command line

h3 login

Log in from this machine with a device code.

Starts a device login. The CLI prints a short code and https://human3.ai/cli/authorize; open that page signed in to Human 3.0, enter the code, check where the request came from, the scopes and the expiry, and approve. The CLI then stores a new API key in ~/.config/h3/credentials.json (mode 0600). The key expires after 90 days unless you pass --expires. Under --json the code is printed to stderr as one JSON line so an agent can pass it to you.

h3 login [--name NAME] [--read-only] [--expires never|DAYS] [--no-browser]

--name NAME

Name for the key on your account page (default: this machine's hostname).

--read-only

Ask for a read-only key; `h3 complete` will then be refused.

--expires never|DAYS

Key lifetime: 1–365 days, or never. Default: 90 days.

--no-browser

Do not try to open the approval page in a browser.

Log in with a read and write key named after this machine.

$ h3 login

A 30-day read-only key for an agent.

$ h3 login --read-only --name claude-agent --expires 30

Output with --json:

{
  "logged_in": true,
  "email": "you@example.com",
  "key": { "id": "…", "name": "laptop", "prefix": "sk_h3_AbCd1234", "scopes": ["read", "write"], "expires_at": "2026-12-25T10:00:00.000Z" },
  "credentials_path": "/home/you/.config/h3/credentials.json"
}

h3 logout

Revoke this machine's key and delete the stored credentials.

Revokes the stored key on the server, then deletes ~/.config/h3/credentials.json. A key supplied through H3_API_KEY is not touched; revoke it on your account page.

h3 logout

Revoke and forget this machine's key.

$ h3 logout

Output with --json:

{ "logged_out": true, "revoked": true, "credentials_removed": true }

h3 whoami

Show who the key belongs to and what it can do.

Your account, membership status, and the key in use (name, prefix, scopes, expiry).

h3 whoami

Check which account this machine is logged in as.

$ h3 whoami

Just the email.

$ h3 whoami --json | jq -r .user.email

Output with --json:

{
  "user": { "id": "…", "email": "you@example.com", "name": "You", "member_since": "2026-03-01T00:00:00.000Z" },
  "subscription": { "status": "active", "plan": "core", "active": true },
  "auth": { "type": "api_key", "key": { "id": "…", "name": "laptop", "prefix": "sk_h3_AbCd1234", "scopes": ["read", "write"], "expires_at": null } }
}

h3 progress

Your curriculum progress, overall and per track.

How many library items you have completed, overall and per track, plus the ids you have completed and viewed.

h3 progress

A per-track table.

$ h3 progress

Overall completion percentage.

$ h3 progress --json | jq .percent

Output with --json:

{
  "completed": 12,
  "total": 64,
  "percent": 19,
  "viewed": 20,
  "tracks": [{ "track": "SO1", "name": "Purpose & Meaning", "completed": 3, "total": 5 }],
  "completed_ids": ["SO1-V1"],
  "viewed_ids": ["SO1-V1", "SO1-E1"]
}

h3 score

Your H3 Maturity Model position and scores.

Where you are on the Human 3.0 Maturity Model: the current stage, how many stages are complete, the markers you have answered, and an overall position from 0 to 1. Also returns your saved H3MM self-scores when you have them.

h3 score

Current stage and per-stage markers.

$ h3 score

Just the stage name.

$ h3 score --json | jq .position.stage.name

Output with --json:

{
  "position": {
    "stage": { "order": 2, "name": "Activated" },
    "stages_complete": 1,
    "stages_total": 5,
    "markers_answered": 7,
    "markers_total": 20,
    "position": 0.28
  },
  "stages": [{ "order": 1, "name": "Aware", "answered": 4, "total": 4, "complete": true }],
  "scores": { "SO1": 3 }
}

h3 badges

Badges you have earned and the ones still open.

Every badge with your level in it: 0 unearned, 1 Practicing, 2 Embodying.

h3 badges [--earned]

--earned

Only badges at level 1 or above.

What you have earned so far.

$ h3 badges --earned

Names of badges at Embodying.

$ h3 badges --json | jq '.badges[] | select(.level == 2) | .name'

Output with --json:

{
  "earned": 5,
  "total": 39,
  "badges": [{ "id": "navigator", "name": "Navigator", "description": "You know where you're going", "domain": "SO", "level": 1, "level_name": "Practicing", "earned_at": "2026-09-01 10:00:00" }]
}

h3 library

List the library, optionally one track or type.

Every active item in the library in curriculum order, with whether you have viewed or completed it. Video items link to their page on human3.ai; the API never hands out a playable video URL.

h3 library [--track TRACK] [--type TYPE] [--todo]

--track TRACK

A track (SO1…CA4) or a whole domain (SO, SH, CO, CA).

--type TYPE

Only items of this type (video, exercise, book, essay…).

--todo

Only items you have not completed.

The Purpose & Meaning track.

$ h3 library --track SO1

Everything left in the Capabilities domain, as JSON.

$ h3 library --track CA --todo --json

Output with --json:

{
  "count": 1,
  "items": [
    {
      "id": "SO1-V1",
      "track": "SO1",
      "track_name": "Purpose & Meaning",
      "type": "video",
      "title": "Finding Your TELOS",
      "description": "…",
      "duration_seconds": 1260,
      "completed": false,
      "viewed": true,
      "url": "https://human3.ai/content/SO1-V1",
      "video_page": "https://human3.ai/v/<stream-uid>/iframe"
    }
  ]
}

h3 item

One library item in full.

Everything about one item: description, why it matters, its steps, and links.

h3 item <id>

<id>

The item id, as shown by `h3 library`.

Read the steps for an exercise.

$ h3 item SO1-E1

Output with --json:

{
  "item": {
    "id": "SO1-E1",
    "track": "SO1",
    "track_name": "Purpose & Meaning",
    "type": "exercise",
    "title": "Write your draft mission statement",
    "description": "…",
    "importance": "…",
    "steps": ["…"],
    "duration_seconds": null,
    "completed": false,
    "viewed": false,
    "has_file": false,
    "essay_url": null,
    "url": "https://human3.ai/content/SO1-E1",
    "video_page": null
  }
}

h3 next

The item to do next.

The item you started but have not finished, if there is one; otherwise the first item you have not completed, in curriculum order.

h3 next

What to work on now.

$ h3 next

Mark the next item done (needs a write key).

$ h3 complete "$(h3 next --json | jq -r .item.id)"

Output with --json:

{
  "reason": "in_progress",
  "item": {
  "id": "SO1-V1",
  "track": "SO1",
  "track_name": "Purpose & Meaning",
  "type": "video",
  "title": "Finding Your TELOS",
  "description": "…",
  "duration_seconds": 1260,
  "completed": false,
  "viewed": true,
  "url": "https://human3.ai/content/SO1-V1",
  "video_page": "https://human3.ai/v/<stream-uid>/iframe"
}
}

h3 complete

Mark an item complete.

Marks the item complete, re-evaluates your badges exactly as the web does, and reports any badge that changed. Needs a key with the write scope.

h3 complete <id>

<id>

The item id.

Mark the mission-statement exercise done.

$ h3 complete SO1-E1

Output with --json:

{
  "item_id": "SO1-E1",
  "completed": true,
  "already_completed": false,
  "badges_changed": [{ "id": "navigator", "name": "Navigator", "level": 1, "level_name": "Practicing" }],
  "progress": { "completed": 13, "total": 64, "percent": 20 }
}

h3 courses

The individual courses for sale.

The à-la-carte course catalog, newest first. Every one is already included in the membership.

h3 courses

List the courses and prices.

$ h3 courses

Output with --json:

{
  "courses": [{ "id": "…", "title": "…", "kind": "video", "tier": "video-tier-1", "description": "…", "price_cents": 4995, "price": "$49.95", "duration_minutes": 83, "published_at": "2026-07-21", "url": "https://human3.ai/courses/…" }]
}

h3 usage

How much of your rate limits you have used.

Your key's per-minute budget, your per-minute budget across all your keys, and your per-day quota: limit, used, remaining, and seconds until each resets.

h3 usage

Check headroom before a batch job.

$ h3 usage

Output with --json:

{
  "key_minute": { "limit": 120, "used": 3, "remaining": 117, "reset_seconds": 41, "window_seconds": 60 },
  "member_minute": { "limit": 240, "used": 3, "remaining": 237, "reset_seconds": 41, "window_seconds": 60 },
  "member_day": { "limit": 10000, "used": 210, "remaining": 9790, "reset_seconds": 51000, "window_seconds": 86400 }
}

h3 completion

Print a shell completion script.

Prints a completion script for bash, zsh or fish. Add it to your shell's startup file.

h3 completion <bash|zsh|fish>

<shell>

bash, zsh or fish.

Install zsh completion (with ~/.zfunc on your fpath).

$ h3 completion zsh > ~/.zfunc/_h3

Install bash completion.

$ echo 'eval "$(h3 completion bash)"' >> ~/.bashrc

Output with --json:

{ "shell": "zsh", "script": "#compdef h3 …" }

h3 version

Print the CLI version.

The CLI version and the runtime it is running under.

h3 version

Which version is installed.

$ h3 version

Output with --json:

{ "version": "1.0.0", "runtime": "node", "runtime_version": "22.9.0" }

h3 help

Show help for all commands or one.

Lists every command. With --json, prints this whole catalog (commands, flags, exit codes, environment, endpoints) for an agent to read.

h3 help [command]

<command>

Show help for this command only. Optional.

Help for one command.

$ h3 help library

The machine-readable catalog.

$ h3 help --json

Output with --json:

{ "name": "h3", "version": "1.0.0", "commands": [ … ], "global_flags": [ … ], "exit_codes": [ … ], "env": [ … ], "endpoints": [ … ] }