feat: joy-check 100% threshold + deterministic routing scripts + CI#633
Closed
notque wants to merge 4 commits into
Closed
feat: joy-check 100% threshold + deterministic routing scripts + CI#633notque wants to merge 4 commits into
notque wants to merge 4 commits into
Conversation
- SKILL.md: pass criteria changed from >= 60 to == 100 (both modes); --strict flag threshold updated to 100; subtle-pattern instruction reframed positively - instruction-rubric.md: pass criteria updated to score == 100 AND zero primary negative patterns - writing-rubric.md: pass criteria updated to score == 100 AND zero GRIEVANCE paragraphs Joy-check instruction-mode analysis of SKILL.md: 100 — zero primary negative patterns. All negative-word occurrences are inside code blocks (contextual exceptions).
Extends validate_positive_instruction_docs.py with the two missing primary patterns (NEVER caps, Don't instruction-start) from the instruction rubric, tightens the Avoid heading regex to eliminate a false positive on technical phrases like "to Avoid N+1", and adds voice-corpus files to the allowlist as a documented contextual exception. New test file covers all 7 primary patterns as golden fixtures, 5 contextual exceptions that must pass (fenced blocks, blockquotes, subordinate lowercase never, clean files, positive rewrites), and a parametrized fleet scan across 44 agents and 116 SKILL.md files: 2 voice corpus files skipped (allowlisted), 177 other components pass. Adds a `joy-check` CI job to test.yml that runs on every push/PR.
- do-classify.py: request classification (complexity, creation, interview, parallel) - do-enhance.py: enhancement stacking (anti-rat, thinking, model selection) - do-build-prompt.py: prompt templates (haiku, banner, agent, task-spec) - do_b SKILL.md: /do with script-backed tables (56% token reduction) - test_do_routing.py: 20 golden routing cases (40 tests) - test_do_enhancement_stacking.py: 17 golden enhancement cases (28 tests) All 247 tests pass across routing, enhancement, and joy-check suites.
Parenthesized assert messages collapsed to single-line f-strings per ruff format rules (line length fits within 120 char limit).
Owner
Author
|
Closing: mixed joy-check + do_b experimental work. Will re-submit joy-check only as a clean PR after resolving private→public skill migration. |
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
1 participant
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.
Summary
do-classify.py,do-enhance.py,do-build-prompt.py— move lookup tables and templates out of /do SKILL.md into testable Python (56% token reduction in do_b variant).validate_positive_instruction_docs.py.Changes
joy-check/SKILL.mdjoy-check/references/instruction-rubric.mdjoy-check/references/writing-rubric.mdscripts/do-classify.pyscripts/do-enhance.pyscripts/do-build-prompt.pyscripts/tests/test_do_routing.pyscripts/tests/test_do_enhancement_stacking.pyscripts/tests/test_joy_check_instruction_mode.pyscripts/validate_positive_instruction_docs.py.github/workflows/test.ymlskills/meta/do_b/SKILL.mdTest plan