Skip to content

Civil AI v1.2.0 - adaptive learning preview

Latest
Latest

Choose a tag to compare

View all tags
@zhangl1001 zhangl1001 released this 11 Aug 04:26
· 10 commits to main since this release
v1.2.0
e91469d
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Civil AI v1.2.0 - adaptive learning preview

Civil AI v1.2.0 strengthens the personal learning loop and makes the public project easier to install, evaluate, and contribute to.

Highlights

  • Daily plans now allocate learning, practice, review, and accumulation work from the candidate's exam cycle and current learning evidence.
  • Daily knowledge, lectures, essay prompts, monthly reviews, and objective questions use bounded recent-content context to reduce repeated output.
  • Duplicate detection can request one alternative teaching angle without turning content diversity into an unbounded retry loop.
  • Grading and error diagnosis remain deterministic and evidence-driven.
  • Public roadmap, contribution workflow, Issue templates, changelog, and upgrade documentation are now part of the repository.

Download

  • civil-ai-web-v1.2.0.zip: browser-ready production bundle. Extract and serve it through a static HTTP server.
  • SHA256SUMS.txt: SHA-256 checksum for the downloadable bundle.

An iOS IPA is not published publicly because development signing is tied to an Apple Developer Team and registered devices. Build the iOS application with your own signing identity using the documented Xcode workflow.

Upgrade

No destructive migration is expected. Export a local backup before upgrading and review the v1.2.0 upgrade guide.

Verification

The release candidate must pass:

  • locked dependency installation and production dependency audit;
  • architecture, database, provider, generation, enrichment, agent, planning, and learning-policy verifiers;
  • TypeScript type checking and Vite production build;
  • design-system checks and mobile smoke verification.

Known limitations

  • The project remains a development preview and still needs broader physical-device coverage.
  • Users provide their own supported model-provider credentials.
  • No third-party or copyrighted question bank is bundled.
  • AI-generated learning content is not official exam guidance and should be reviewed critically.

See the changelog, public roadmap, security policy, and content policy for details.

Assets 4
Loading