feature-request: distribute via Homebrew tap and/or npx-style installer (lower install friction)

[pstuart]

Why

Step 5e first competitive pass (autoscan 2026-05-08) — the single largest competitive gap vs the dominant Claude Code statuslines:

Competitor	Stars	Install command
ccstatusline	~8.9k	npx -y ccstatusline@latest
CCometixLine	~2.9k	npm i -g @cometix/ccline (Rust binary)
claude-powerline	~1.1k	npm i -g @owloops/claude-powerline
cc-statusline (Barista's stated inspiration)	~592	npx install
Barista	(current)	git clone && ./install.sh

A statusline is a try-and-discard tool — users compare options at the command line. The friction gap of git clone + interactive installer vs a single npx -y one-liner is real and likely the largest barrier to adoption.

Key constraint to preserve: Barista's "no Node runtime needed" positioning is a genuine differentiator (every npm-based competitor requires Node). A Homebrew tap satisfies the zero-runtime story while removing the clone friction.

What

Two complementary distribution paths (either or both):

Option A — Homebrew tap (recommended primary path)

• New repo pstuart/homebrew-barista (or fold into pstuart/homebrew-tap if a fleet tap exists)
• Formula: standard Class Barista < Formula with url, sha256, install block running the equivalent of install.sh's file copying
• User installs via brew install pstuart/barista/barista (or shorter if folded into a fleet tap)
• Update on brew upgrade — replaces the in-installer update checker for Brew users

Option B — npx-friendly wrapper (secondary path)

• Optional. A tiny npm package that shells out to download Barista's release tarball and run install.sh
• Lower priority because it betrays the no-Node story; only ship if someone asks for it
• Skip unless adoption stalls

Acceptance (Option A)

• pstuart/homebrew-barista (or shared tap) created with Barista formula
• Formula installs all required modules + barista.sh + barista.conf default into $(brew --prefix)/opt/barista/
• Generates a stub ~/.config/claude-code/statusLine snippet pointing at the brewed path on first install (or via barista init post-install)
• README "Install" section adds brew install pstuart/barista/barista as the primary path; git clone && ./install.sh retained as the from-source path
• CI / release workflow generates the formula's sha256 automatically when a new version tag is cut (single GitHub Actions / shell-script step — out of scope for this issue if CI policy restricts; could be a manual update for now)
• Smoke-test: brew install pstuart/barista/barista && claude shows the Barista statusline correctly

Tier

Vertical-slice eligible Phase 1 — Homebrew formula + tap repo. ≤5 files. Internally complete (brew install works), reversible (delete tap repo + remove formula reference). Phase 2 (npm wrapper) is a separate decision after Brew adoption signal.

Priority: HIGH — single largest competitive-parity gap; npm/bunx is the de-facto distribution channel for Claude-adjacent CLI tooling.

Smallest unblock ask

Pamela / Patrick's call on whether to use a dedicated tap repo pstuart/homebrew-barista or fold into an existing fleet tap (e.g., a hypothetical pstuart/homebrew-tap shared across the Better-* family of CLI tools).

Source

Filed by autonomous scanner — Step 5e first competitive pass (2026-05-08). GAP-1 of 7 surfaced; ranked High by adoption-friction analysis. Competitor evidence: ccstatusline, CCometixLine, claude-powerline.

[pstuart]

9-day re-ping (autonomous scan 2026-05-17)

9.61 days author-silent, past the dossier's 5-day re-ping threshold by 4 days. Still the single smallest unblock ask is:

Pamela / Patrick's call on whether to use a dedicated tap repo pstuart/homebrew-barista or fold into an existing fleet tap (e.g., a hypothetical pstuart/homebrew-tap shared across the Better-* family of CLI tools).

Why this is still HIGH priority

• The competitive-gap analysis from the original filing (2026-05-08) hasn't moved — ccstatusline / CCometixLine / claude-powerline / cc-statusline all still ship one-liner installers.
• Barista's "no Node runtime" positioning makes Homebrew the natural primary path (npm-free).
• Phase 1 (tap repo + formula) is genuinely ≤5 files of new code; the bottleneck is the tap-naming decision, not the engineering.

Re-ping recipe applied per dossier protocol

This is the first re-ping; per scanner protocol the next escalation tier (~14-day mark, around 2026-05-22) is a refined scoping note proposing a default (e.g. "defaulting to pstuart/homebrew-barista dedicated tap unless a fleet-tap convention exists by 2026-05-22"). For now, the ask remains unchanged — just naming.

If a fleet-tap decision is already documented somewhere (CLAUDE.md, internal note, Better-* convention), pointing this issue at it would unblock immediately.

Companion issue #24 (runtime TUI) re-pinged in parallel under the same 9-day-mark protocol.