We garnish our pies. It seemed rude not to extend Pi the same courtesy. This is a Pi extension that adds some quality of life improvements to Pi's UX, while looking good in the process:
User Prompt — high-vis bordered prompt box with configurable border color and style, pi icon toggle, timestamp, and provider/model label. Prompts submitted while Pi is working (steer / Alt+Enter follow-up) are left undecorated so Pi's native Steering: / Follow-up: queue rows still render.
“Working” Loader — animated spinner (with color choice), randomized activity word, optional word packs, shimmer, token activity monitor, elapsed timer, output token display, live output-token rate, and the response model when it meaningfully differs from the selected model — all arrangeable left to right. During a blocking extension prompt, the loader switches to a stable dim waiting line with a pulsing indicator, then resumes afterward.
Completion Marker — end-of-turn marker with icon, randomized verb (or Worked; matching past tense when Randomize "Worked" text is on), token consumption display, and optional marker-style, border-style, and border-color decorations spanning the full terminal width. The default elite style trails the summary with tapering dashes; bookend fills the remaining width to a closing corner glyph. Hooks into Pi's agent_settled event.
pi install npm:@underactive/pi-toppingRestart Pi (or run /reload) to pick it up.
Run /topping-settings (TUI only) to customize your toppings. Settings persist to
~/.pi/agent/pi-topping/settings.json; hand edits are tolerated on load.
╔═[ Pi Topping: Settings ]══════════════════════════════════════╗
╟─ Preview ─────────────────────────────────────────────────────╢
║ ║
║ ⠋ Crafting… ⣠⣤⣶⣶⣤⣠⣀⢀ 28 tps · 3s · ↓ 84 tokens · test-model ║
║ ║
╟─ User Prompt ─────────────────────────────────────────────────╢
║ [■] High-vis prompt ON ║
║ [■] Border style ‹ double › ║
║ [■] Border color ‹ borderAccent › ║
║ [■] Pi icon ON ║
║ [■] Timestamp ON ║
║ [■] Provider ON ║
║ [■] Model ON ║
║ ║
╟─ “Working” Loader ────────────────────────────────────────────╢
║ [■] Animated spinner ON ║
║ ❯ [■] Animated spinner color ‹ accent › ║
║ [■] Randomize “Working” text ON ║
║ [■] Text shimmer ON ║
║ [■] Invert shimmer OFF ║
║ [■] Text shimmer direction ‹ Left to Right › ║
║ [■] Text shimmer speed ‹ Normal › ║
║ [■] Token activity monitor ON ║
║ [■] Token activity monitor color ‹ accent › ║
║ [■] Token activity monitor direction ‹ Right to Left › ║
║ [■] Token activity monitor dimmed OFF ║
║ [■] Elapsed time since prompt ON ║
║ [■] Show output tokens ON ║
║ [■] Token rate ON ║
║ [■] Token rate color ‹ warning › ║
║ [■] Token rate dimmed OFF ║
║ [■] Response model ON ║
║ [■] Response model color ‹ accent › ║
║ [■] Response model dimmed OFF ║
║ ║
╟─ Elements Order ──────────────────────────────────────────────╢
║ [ ] Animated spinner ║
║ [ ] “Working” text ║
║ [ ] Token activity monitor ║
║ [ ] Token rate ║
║ [ ] Elapsed time ║
║ [ ] Output tokens ║
║ [ ] Response model ║
║ ║
╟─ Completion Marker ───────────────────────────────────────────╢
║ [■] Show completion marker ON ║
║ [■] Marker style ‹ elite › ║
║ [■] Border style ‹ none › ║
║ [■] Border color ‹ borderAccent › ║
║ [■] Pi icon ON ║
║ [■] Randomize “Worked” text ON ║
║ [■] Tokens spent ON ║
║ [■] Mid-turn inputs ON ║
║ ║
╟─ Word Packs ──────────────────────────────────────────────────╢
║ [ ] Doctor Who OFF ║
║ [ ] Firefly OFF ║
║ [ ] Hitchhiker's Guide OFF ║
║ [ ] The Lord of the Rings OFF ║
║ [ ] The Matrix OFF ║
║ [ ] Portal OFF ║
║ [ ] SimCity OFF ║
║ [ ] Star Trek OFF ║
║ [ ] Star Wars OFF ║
║ ║
╟─ Options ─────────────────────────────────────────────────────╢
║ [■] Use NerdFont icons ON ║
╟───────────────────────────────────────────────────────────────╢
║ ↑↓ move PgUp/PgDn page ←→ select ║
║ ␣ toggle ⏎ apply esc cancel ║
╚═══════════════════════════════════════════════════════[ 9/50 ]╝
Base activity words are always available. All word packs, including the shipped Doctor Who, Firefly, Hitchhiker's Guide, The Lord of the Rings, The Matrix, Portal, SimCity, Star Trek, and Star Wars packs and custom packs, are disabled by default. Enable packs in /topping-settings to add their entries to the same uniformly selected pool. When Randomize “Worked” text is on, the completion marker uses the selected entry’s matching past tense; otherwise it uses Worked.
Bundled packs live in wordpacks/. Copy wordpacks/doctor-who.json, wordpacks/firefly.json, wordpacks/hitchhikers-guide.json, wordpacks/lord-of-the-rings.json, wordpacks/matrix.json, wordpacks/portal.json, wordpacks/simcity.json, wordpacks/star-trek.json, or wordpacks/star-wars.json to ~/.pi/agent/pi-topping/word-packs.json, then change the pack id (the bundled doctor-who, firefly, hitchhikers-guide, lord-of-the-rings, matrix, portal, simcity, star-trek, and star-wars IDs are reserved), name, and words:
{
"packs": [{
"id": "cooking",
"name": "Cooking",
"words": [{ "present_tense": "Sautéing onions", "past_tense": "Sautéed" }]
}]
}Pack IDs must start with a lowercase letter and contain only lowercase letters, digits, and -. Each pack needs a non-empty name and at least one word with non-empty present_tense and past_tense strings. Invalid entries are ignored. Packs reload when a session starts and whenever /topping-settings opens, not while a turn is running. Preferences for missing packs are retained so they apply when the pack returns.
Under Elements Order, press ␣ to grab a row, then ↑/↓ to slide that element
left or right within the loader. Elapsed time, output tokens, token rate, and response model are joined with
· separators whenever adjacent. When reordering separates them, each adjacent run is rendered
without a separator before or after it.
The N tps token rate uses the selected theme color and
remains reorderable. Active rates below 1,000 are padded to three characters so updates do not shift the other
segments; larger rates may shift later segments. When inactive, it remains as the dim --- tps placeholder. After its last update it holds
full brightness for 1.5 seconds, then fades through five theme-aware shades to the dim text color over
the next 0.25 seconds before returning to the placeholder; a new count restores full brightness and
restarts the cycle.
The response model is captured from assistant responses, displayed as its sanitized value with no label only when it meaningfully differs from the selected model, and defaults to the final loader detail after output tokens. Equivalent local paths and decorated filenames are suppressed. It can be hidden, recolored, permanently dimmed, or reordered. After the agent settles, its final value remains visible for 3 seconds, fades through five theme-aware shades over 0.5 seconds, then clears; new work cancels the pending fade.
- Pi. Pi ≥0.84.4 is required for the waiting-for-input loader state; older Pi versions keep the animated loader.
- Node.js ≥22.19.0
npm install
npm test
npm run typecheckTests cover the word/shimmer rendering, prompt box and completion marker rendering, activity meter behavior, timer resets, settings persistence, the menu.ts component, and the settings menu wiring end to end.
- Token counts are a streaming word-count estimate, not an exact tokenizer count — they self-correct when each message finishes, so you'll see a small jump at that point.
- Elapsed time normally starts when you submit a prompt; for auto-continuations without that event, it starts slightly later instead.
- If another extension also customizes the working message, whichever one runs last wins.
Eric Sison