spec(crm-story-reel): v23 -- house marks read as a reader reads them (#5897) - #5898
Merged
Merged
Conversation
…5897) The text and every house mark are folded the same way first: NFKC, invisible characters dropped, one case, each Cyrillic look-alike read as its Latin letter (FOLD_LOOKALIKES, LOOKALIKE_LATIN, INVISIBLES_DROPPED), so a Latin T inside the Russian spelling of the name is the house again. code_mark: T27 starting a word is the brand (t27ai, t27_ai), not after a letter or digit (at27) nor before a digit (T270). product_mark: S3AI is one code token; "our S3 AI pipeline" names no house. name_mark: "Trinity College" stays a house name on purpose -- no rule on the next word clears it without clearing "Trinity AI"; a refused agent line is rewritten for free, a missed one is on her reel. 60/60 spec tests, vacuity 0 of 60, 93/93 negative controls caught (9 new for v23); seal verified; dupe_scan: no new duplicate body. Closes #5897 Refs gHashTag/999-multibots-telegraf#3564 phi^2 + 1/phi^2 = 3 | TRINITY Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
Owner
Author
Reviewer bee: approved (spec 0887fd3 / host 43ec939cb)I ran my own probes with
Per-line test (
Gates
CI
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
crm-story-reel v23 restates HOUSE_MARKS: the detector that crm-story-reel v22 and split-reel v7 read on every line of a client's reel (999
render/src/house-names.ts, gHashTag/999-multibots-telegraf#3564).write to t27aiour S3 AI pipelineI studied at Trinity CollegeRules
INVISIBLES_DROPPED), one case, each Cyrillic look-alike read as its Latin letter (FOLD_LOOKALIKES,LOOKALIKE_LATIN). One canonical form, so both directions fold.CODE_MARK_PREFIX: t27, t27ai, t27_ai, T27.ai), not after a letter or digit (at27, 5t27) nor before a digit (T270).PRODUCT_ONE_TOKEN): S3AI and S3AIs are the house;S3 AIwith a space is S3 the storage and the word AI.NAME_INSIDE_WORD: TrinityBot) and before any word (NAME_IN_ANY_CONTEXT). Trinity College stays flagged: no rule on the next word clears the college without also clearing "Trinity AI" and "Write To Trinity AI". The costs are asymmetric: a refused agent line is rewritten for free, while a missed one puts the house's name on a client's reel. Her own typed line is still hers (own_words).Checks (from command output)
t27c test-report: 60/60 passt27c validate-vacuity: 0 of 60 vacuoustri spec-controls(999 binding): 93 of 93 caught, 9 of them new for v23t27c seal --verify: all hashes MATCHtools/dupe_scan.py: no new duplicate body, and none of the known groups grewHost binding: gHashTag/999-multibots-telegraf, branch
fix/house-names-v23.Closes #5897
🤖 Generated with Claude Code