Repository navigation
feat(fp): chat answers in context, named mayors reply in person - #14
Merged
Merged
Conversation
Replies in the global chat and face to face now answer what the player
asked with what is really going on. scripts/fp/chat/chat_facts.gd puts
the district mayor's last turn in words ("zoned 6 homes near Elm Street
and ordered a fire station"), their post, the Gazette headline, the
council's last ruling, incidents, sites going up, the district's numbers
and its top complaints into the player-reply and talk prompts as
city_facts, and the offline answers quote the same facts.
A mayor the player names, or the district's mayor for a bare "mayor",
answers first and in person: their card leads the live prompt with their
last turn, and when the live reply leaves them out (or offline) they
answer from their facts.
Replies to the player are no longer dropped as repeats of earlier rows,
each reply call gets its own seed, and the drip holds everyday lines
while the player is being answered. A typing row above the input names
who answers next, and a Chat status card says when live replies fail
twice in a row and when they come back. A new chat_reply photo-tour shot
shows a question to a mayor being answered.
fp_chat_test covers the live and offline answers, the typing row and the
status card; the new fp_chat_context_test covers the facts in words,
mayor names, forced replies, the drip hold and the typing line.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Automated reviewNo new findings on the 18 changed files (gdlint, complexity, TODO policy, dead code). Run |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Fixes "the chat doesn't reply in context" from the engagement feedback. Replies in the global chat and in face-to-face talks now answer what the player asked, using what is really going on in the city.
scripts/fp/chat/chat_facts.gd, no nodes):city_facts, and both prompts now ask the model to answer what the visitor asked from them./mayordecrees keep the "To the mayors:" prefix in the reply call, so the district's mayor answers them.chat_replyshot asks the district's mayor about their day and catches the answer while the next neighbour is typing.chatfeature.Checks
GATE_RETRIES=1 tools/gate.sh fullon d251a13: PASS, 33 of 33 steps on the first attempt (about 5 minutes, no retries).pre-commit run --all-filesandpython tools/checks/run_all.py: PASS. API docs regenerated.ChatFeed.fast_countleaves the dead-code baseline, because the typing row uses it now.fp_chat_testadds the live and offline answers in context, the typing row, the drip hold, the status card and the tour hook. The newfp_chat_context_testcovers orders and rulings in words, mayor names, forced replies, the hold and the typing line.chat_reply, Mayor Kimura answers with her real last turn, a plumber follows, and the typing row names the next neighbour.