buhtig 0.1.0-alpha.3 → 0.1.0-alpha.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -387,7 +387,7 @@ A repo can override the *choice* (`agent: codex` in its own entry, or in `.git/b
387
387
  Until something chooses, the Alt+W → Agent action reads "configure…" and takes you to the settings
388
388
  page rather than launching whatever buhtig happened to be written against.
389
389
 
390
- A launched agent is **briefed on the way in**: it is told that you can send it a one-line instruction
390
+ A launched agent is **briefed on the way in**: it is told that you can send it a one-line message
391
391
  from the UI, and given one exact command to watch for one — `buhtig agent listen`, pointed at
392
392
  `$BUHTIG_STREAM` and chosen for the agent being launched (Claude Code sets it up as a persistent
393
393
  monitor that never returns and wakes it on each message; Codex and opencode run it for one bounded
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "buhtig",
3
- "version": "0.1.0-alpha.3",
3
+ "version": "0.1.0-alpha.4",
4
4
  "description": "Local-first, keyboard-driven pull-request review cockpit",
5
5
  "license": "UNLICENSED",
6
6
  "type": "module",
@@ -246,8 +246,11 @@ For each:
246
246
  answered with `review.updateRun` rather than an edit to any finding.
247
247
  - **Oldest first.** A later message often refines an earlier one, and answering the refinement first
248
248
  produces an edit the earlier one then overwrites.
249
- - **The edit is the answer.** The reply is one line saying what you changed, or why you did not — it is
250
- not a place to argue, and there is no thread for the human to reply into.
249
+ - **The message says whether to edit.** "Rewrite the title" and "does this still hold?" arrive through
250
+ the same box and look identical until you read them. Where it asks for a change the change is the
251
+ answer, and the reply is one line saying what you changed or why you did not; where it asks a
252
+ question, the reply is the answer. Neither is a place to argue, and there is no thread for the human
253
+ to reply into.
251
254
  - **Disposition only when the message says so.** Several of the canned prompts do — "verify" and "already
252
255
  handled?" both end with "dismiss it as `not-applicable` if it does not hold", and that is an
253
256
  instruction to carry out rather than a claim to argue with. A free-text "this is nonsense" is not: