vesta

A small Python formatter / previewer / publisher for Vestaboard devices.

semantic input → board layout → terminal preview → optional publish

What it does

• Formats structured data (JSON, CSV, plain text) for Vestaboard
• Supports multiple device profiles:
  • flagship: 6 × 22
  • note: 3 × 15
• Previews output in the terminal before sending
• Publishes via Vestaboard Cloud API or Local API

Templates

metrics and table are accepted as aliases for data.

CSV is auto-detected — no --template flag needed.

Key suffixes

Field name suffixes control formatting automatically. The suffix is stripped from the label on the board.

{
  "revenue_curr": 84210.50,
  "sessions": 10823,
  "growth_pct": 12.4
}

Renders as:

REVENUE   $84.2K
SESSIONS  10.8K
GROWTH    12.4% ██

Color indicators

Color is driven by semantic tone, not raw cell placement. The trailing colored tile is added automatically when a tone can be determined.

Auto-detection: inferred when a field name contains pct, percent, change, delta, or diff and the value is numeric:

• positive → green
• negative → red
• zero → white

Explicit tone: set via _style:

{
  "score": 91.2,
  "_style": { "score": "good" }
}

Accepted tone names: good, bad, warn, info, neutral, muted, or a direct color: green, red, yellow, blue, white, black, violet, orange.

Range-based tone: specify good and bad thresholds for a 4-step gradient. Direction is implicit — wherever good sits numerically is the green end:

{
  "bounce_rate": 68.4,
  "_style": { "bounce_rate": {"good": 30, "bad": 80} }
}

_style and other _-prefixed keys are never shown on the board.

Use --explain to see which fields got indicators and why:

cat metrics.json | vesta render --template data --preview-only --explain

Layout flags

--title TEXT

Adds a title row at the top with colored tile bookends. Tries 2 tiles each side, falls back to 1 if the text is long.

--title-color COLOR[,COLOR,COLOR]|none

Color of the bookend tiles. Defaults to white. Pass none for a plain centered title with no tiles. Pass up to 3 comma-separated colors to use multiple tiles — the right side mirrors the left: --title-color red,blue,orange places red blue orange on the left and orange blue red on the right. Falls back through fewer tiles if the title text is too long to fit.

--subtitle TEXT|time

Optional second row below the title, with a single tile bookend on each side (same color as title). Use the special value time to insert the current time. You can also embed the subtitle directly in the title using a newline — --title $'Weather\nSan Francisco' — and the second line becomes the subtitle automatically (explicit --subtitle takes precedence).

--separator [PATTERN]

Adds a full-width row of colored tiles below the title block (after subtitle if present). Used alone, defaults to solid white. Accepts:

echo '{"temp": "72F", "wind": "12mph", "humidity": "45%"}' | \
  vesta render --template kv --title "Weather" --separator rainbow --preview-only

┌────────────── flagship 6x22 ───────────────┐
│████          W E A T H E R             ████│
│████████████████████████████████████████████│
│T E M P                               7 2 F │
│W I N D                           1 2 M P H │
│H U M I D I T Y                       4 5 % │
│                                            │
└────────────────────────────────────────────┘

vesta render --input weather.json --template kv --columns 2 \
  --title "BROOKLYN WEATHER" --title-color white --subtitle time --preview-only

┌────────────── flagship 6x22 ───────────────┐
│████  B R O O K L Y N   W E A T H E R   ████│
│██              1 0 : 4 3 P               ██│
│N O W     6 2 F     R A I N           0 %   │
│L I K E   6 2 F     W I N D     1 5 M P H   │
│H I G H   6 6 F     U V             4 . 4 ██│
│L O W     4 8 F     S E T     7 : 3 2 P M   │
└────────────────────────────────────────────┘

--columns [1|2] (kv template only)

Pack two key-value pairs per row instead of one. Each column is sized independently to its own content, which creates a natural gap between columns. Color indicators from _style or auto-detection still apply: left-column tiles appear in the gap; right-column tiles appear at the board's right edge.

echo '{"now":"62F","rain":"0%","high":"66F","low":"48F"}' | \
  vesta render --template kv --columns 2 --preview-only

┌────────────── flagship 6x22 ───────────────┐
│N O W     6 2 F             R A I N     0 % │
│H I G H   6 6 F             L O W     4 8 F │
│                                            │
│                                            │
│                                            │
│                                            │
└────────────────────────────────────────────┘

Falls back to --columns 1 with a warning if the content is too wide for the profile.

--align [left|center|right]

For metrics (JSON object): aligns the content block horizontally as a unit. Default is left.

For tables (JSON array or CSV): default is center (compact block, centered). left and right spread columns edge-to-edge with equal inter-column gaps — first column anchored to the chosen edge, last column anchored to the opposite edge.

--valign [top|center]

Vertical alignment of the content block. Default is top. Use center for breathing room when you have fewer rows than the board height.

--timestamp

Adds the current time (10:01A, 9:30P) to the bottom-right corner. Silently skipped if there isn't room. Use --force-timestamp to place it regardless, overwriting content if needed.

--tz

IANA timezone for the timestamp, e.g. America/New_York. Defaults to local system time.

--profile [flagship|note]

Board profile. Auto-detected from API grid dimensions when publishing. Defaults to flagship.

Example usage

Text:

echo '"hello world"' | vesta render --preview-only

┌────────────── flagship 6x22 ───────────────┐
│                                            │
│                                            │
│          H E L L O   W O R L D             │
│                                            │
│                                            │
│                                            │
└────────────────────────────────────────────┘

Key/value:

echo '{"temp": "72F", "wind": "12mph"}' | vesta render --template kv --preview-only

┌────────────── flagship 6x22 ───────────────┐
│T E M P                               7 2 F │
│W I N D                           1 2 M P H │
│                                            │
│                                            │
│                                            │
│                                            │
└────────────────────────────────────────────┘

Key/value 2-col with title:

vesta render --input testdata/weather.json --template kv --columns 2 \
  --title "BROOKLYN WEATHER" --subtitle time --preview-only

┌────────────── flagship 6x22 ───────────────┐
│████  B R O O K L Y N   W E A T H E R   ████│
│██              1 0 : 4 3 P               ██│
│N O W     6 2 F     R A I N           0 %   │
│L I K E   6 2 F     W I N D     1 5 M P H   │
│H I G H   6 6 F     U V             4 . 4 ██│
│L O W     4 8 F     S E T     7 : 3 2 P M   │
└────────────────────────────────────────────┘

CSV table (auto-detected, centered by default):

vesta render --input scores.csv --preview-only

┌────────────── flagship 6x22 ───────────────┐
│    N A M E       S C O R E     R A N K     │
│    A L I C E           9 8           1     │
│    B O B               8 7           2     │
│    C A R O L           7 6           3     │
│    D A V E             6 1           4     │
│                                            │
└────────────────────────────────────────────┘

Use --align left or --align right to spread columns edge-to-edge instead.

Metrics with color indicators:

echo '{
  "revenue_curr": 84210.50,
  "sessions": 10823,
  "conversion_pct": 3.2,
  "bounce_rate_pct": 68.4,
  "_style": {
    "revenue_curr": "good",
    "conversion_pct": {"good": 8, "bad": 2},
    "bounce_rate_pct": {"good": 30, "bad": 80}
  }
}' | vesta render --template data --valign center --align center --timestamp --preview-only

┌────────────── flagship 6x22 ───────────────┐
│                                            │
│    R E V E N U E   $ 8 4 . 2 K ██          │
│    S E S S I O N S   1 0 . 8 K             │
│    C O N V E R S I O N   3 . 2 % ██        │
│    B O U N C E   R A T E   6 8 . 4 % ██    │
│                                  9 : 3 4 P │
└────────────────────────────────────────────┘

Note profile:

vesta render --input data.json --profile note --template data --preview-only

┌───────── note 3x15 ──────────┐
│T E M P           $ 7 2 . 0 0 │
│H U M I D I T Y       5 4 % ██│
│C H A N G E       - 2 . 1 % ██│
└──────────────────────────────┘

Get raw character codes (for direct API use):

cat data.json | vesta render --json-only

Publishing

Cloud API:

cat data.json | vesta post-cloud --token $VESTABOARD_TOKEN

Local API:

cat data.json | vesta post-local --api-key $VESTABOARD_LOCAL_API_KEY

Preview current board state:

vesta read-cloud

VESTABOARD_TOKEN is read from the environment. Board profile is auto-detected from the grid dimensions returned by the API. Pass --profile to override.

Re-render a saved board:

cat data.json | vesta render --json-only > saved.json
cat saved.json | vesta render --preview-only

Running the examples

./run_examples.sh

Runs all bundled examples against local test data. Requires no API credentials.

Installation

pip install vesta

Or run directly from source with uv:

uv run vesta.py render

Why this exists

Hitting the Vestaboard API directly is straightforward. The harder part is making structured data fit well on a small fixed-size grid — compacting numbers, handling suffixes, previewing locally, and reusing layouts across scripts and data sources. This project is mainly that rendering layer.