Idea: voice preset discovery (list voice_styles after bootstrap)

[omarelkhal]

Problem

After install.sh / bootstrap, new users may not know which voice_type (M1, F4, …) or display name (Sara, James, …) to pick in speak_summary.toml.

Already in the tree (narrow the PR scope)

• /aftertone-voice (Cursor) and /aftertone_voice (Claude) — interactive picker with human labels via py/voice_presets.py
• uv run --directory py python speak_summary_config.py voices — lists presets from assets/voice_styles/*.json
• Daemon status shows current voice (e.g. Sara (female))

Still useful (this issue)

• Print preset list at end of install.sh / bootstrap.sh (one screen, no extra commands)
• README or docs snippet: “where voice JSON lives” + link to slash command
• Optional: sample speak_summary.toml comment block listing ids

Outcome

PR that improves discoverability only — no change to default voice (F4 / Sara) unless discussed separately.

Acceptance

• New contributor can find voices without reading Python source
• Docs mention picker + CLI list command
• No regression to daemon load or TOML parsing

Related: Supertone/supertonic-3 assets under assets/voice_styles/; see README Control.

[omarelkhal]

Scoped down: voice picker + speak_summary_config.py voices already exist — this issue is now about post-install discoverability (install.sh banner, docs snippet).