Skip to content

Claude/review readme governance k ik ow#1

Merged
RegardsKiki2 merged 3 commits intomainfrom
claude/review-readme-governance-KIkOw
Mar 26, 2026
Merged

Claude/review readme governance k ik ow#1
RegardsKiki2 merged 3 commits intomainfrom
claude/review-readme-governance-KIkOw

Conversation

@RegardsKiki2
Copy link
Copy Markdown
Contributor

No description provided.

claude added 3 commits March 26, 2026 19:59
- Create comprehensive README with app documentation, setup guide,
  architecture overview, and a section teaching about governance files
- Add CONTRIBUTING.md with contribution guidelines and dev setup
- Add CODE_OF_CONDUCT.md based on Contributor Covenant 2.1
- Add SECURITY.md with vulnerability reporting and security architecture
- Fix package.json license field from "MIT" to "Apache-2.0" to match
  the actual LICENSE file
- Update LICENSE copyright from placeholder to "2025 NeuroverseOS"

https://claude.ai/code/session_01Fo4Up2mMrkuQv7bUS1PiqG
- Add tsconfig.json for deterministic TypeScript builds (ES2022, NodeNext)
- Add vitest with 78 tests across 3 test files:
  - governance.test.ts: gate transitions, word limits, trust degradation chain
  - sign-loader.test.ts: world loading, parsing, stacking, compatibility
  - patterns.test.ts: speech recognition regex validation
- Extract governance logic into testable src/governance.ts module
- Add type declarations for private platform dependencies (external.d.ts)
- Add GitHub Actions CI workflow (typecheck, test, docker build)
- Add .gitignore for node_modules and dist
- Fix server.ts import path for NodeNext module resolution

https://claude.ai/code/session_01Fo4Up2mMrkuQv7bUS1PiqG
StarTalk must always interpret meaning through symbolic systems (astrology)
and must never output tactical advice. This is the product boundary —
StarTalk is a translator, not a coach.

Adds validateCoreInvariant() which checks two rules:
- hasSymbolic: response references astrological concepts (signs, elements,
  planets, modalities, traits, shadow, energy, chart)
- hasTactical: response contains concrete directives ("you should buy",
  "step 1:", "action items:", "here's a plan")

Both must hold: symbolic AND not tactical. Astrology-garnished tactical
advice still fails. Generic non-symbolic responses also fail.

15 new tests covering valid symbolic reads, invalid tactical advice,
tactical-wrapped-in-astrology, and missing-symbolic-framing cases.

https://claude.ai/code/session_01Fo4Up2mMrkuQv7bUS1PiqG
@RegardsKiki2 RegardsKiki2 merged commit 33cf5dc into main Mar 26, 2026
1 check failed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants