murmur 0.2.1

Twelve commits since v0.2.0 — the radio now tells you when a newer one
is out, voice setup finishes from one pasted key with two murmur voices
ready to clone into your own account, the guide gets a real composer and
sets the speaking rate when you ask, the clock names the day and the beat
speaks from the program actually on air, plus the setup conversation
reading like one and the TUI's missing-piece fix.

What's Changed

• chore(release): the version number is the listener's call, not the radio's by @wine-fall in #182
• feat(commands): the radio checks npm for a newer murmur [spec 10] by @wine-fall in #183
• chore(dev): make pack — rehearse the install a listener actually gets by @wine-fall in #184
• fix(tui): the setup conversation reads like a conversation [spec 10] by @wine-fall in #185
• feat(guide): one pasted key finishes the whole voice, timbre included [spec 03-03] by @wine-fall in #186
• feat(tui): the guide gets a real composer, and your line is in the log at once [spec 10] by @wine-fall in #187
• feat(guide): two murmur voices to clone into your own account [spec 03-03] by @wine-fall in #188
• docs(readme): the logo follows the reader's theme by @wine-fall in #189
• fix(tui): a missing piece is ember, not one more quiet hint [spec 10] by @wine-fall in #190
• fix(director): the beat speaks from the real program, not an imagined one [spec 04] by @wine-fall in #165
• feat(director): the clock names the day, not just the hour [spec 04] by @wine-fall in #191
• feat(voice): speaking rate, set by talking to the guide [spec 02] by @wine-fall in #192
• chore(release): 0.2.1 by @wine-fall in #193

Full Changelog: v0.2.0...v0.2.1