ATTI

ATTI is an Edge extension for supported personality-assessment websites. It keeps profile and session data local by default, uses a provider only when the user explicitly starts planning, and can preview plus fill supported questionnaire pages without auto-submitting them.

Current repository version: 0.5.0

Chinese README: README.zh-CN.md

Current support status:

• highest-confidence live path: Truity Enneagram
• additional adapter-scoped supported routes: Truity DISC, Truity TypeFinder
• narrow second-site sample: 16Personalities, with fixture-backed and browser-level extension coverage but no stable live-site verification in this environment because Cloudflare blocks access
• additional adapter-scoped public test route: SBTI / https://sbti.cc/test, with dedicated extraction/fill support for its single-question stepping flow during explicit fill
• experimental last-resort path: generic fallback extraction/preview for unsupported assessment pages; this remains limited, does not allow page fill by default, and should not be treated as universal website support

Supported Pages

• Truity / Enneagram Personality Test
  • https://www.truity.com/test/enneagram-personality-test
• Truity / DISC Personality Test
  • https://www.truity.com/test/disc-personality-test
• Truity / TypeFinder Personality Test
  • https://www.truity.com/test/type-finder-personality-test-new
• 16Personalities / Free Personality Test
  • https://www.16personalities.com/free-personality-test
• SBTI / test
  • https://sbti.cc/test

Simple Profile Draft Format

For a quick first run, paste a short local profile draft into the side panel using this format.

Profile Summary

I usually prefer calm, cooperative, stable choices. When answering questionnaires, I lean toward responses that are consistent, moderate, and practical rather than extreme.

Evidence Notes

- prefers clear structure
- willing to cooperate with others
- dislikes overly aggressive expression
- tends to choose stable and executable options
- usually observes first in unfamiliar situations

Chinese quick-copy version:

• README.zh-CN.md
• docs/guides/zh-CN/minimal-profile-draft.zh-CN.md

Fill Actions

• Run answer planning generates recommendations and refreshes the side-panel preview without filling the page.
• Apply recommended fill applies the currently previewed recommendations to the active supported page without generating a new answer plan.
• SBTI / test now advances through its one-question-at-a-time flow during the explicit fill step until the page reaches a ready-to-submit state.
• ATTI still does not auto-submit the final questionnaire.

Error Recovery

The side panel classifies errors as retryable or permanent:

• Retryable errors (network timeout, provider transient failure) show a retry button so the user can re-attempt the failed action without starting over.
• Permanent errors (missing API key, extension disabled, domain not approved) display the error message without a retry button, since the user needs to fix configuration first.

Data Management

The options page provides data export and cleanup:

• Export sessions: downloads all session records as a timestamped JSON file.
• Export profiles: downloads all profile records as a timestamped JSON file.
• Purge completed sessions: removes all sessions with status completed after a confirmation dialog. This does not delete active sessions or profiles.

Session Status

The session status card shows structured information about the active session:

• siteId: the detected assessment site (e.g. truity-enneagram)
• Question count: how many questions have been extracted
• Recommendation stats: how many questions have AI recommendations vs total answer plans

Quick Start

Core implementation and architecture work should begin with these files:

1. memory-bank/@architecture.md
2. memory-bank/@game-design-document.md
3. software-design-document.md
4. tech-stack.md

Documentation Map

• Canonical architecture memory: memory-bank/@architecture.md
• Product intent and constraints: memory-bank/@game-design-document.md
• Lightweight implementation architecture: software-design-document.md
• Engineering and stack rules: tech-stack.md
• Repository structure guide: docs/repository-map.md
• Documentation maintenance rules: docs/documentation-rules.md
• Versioning rule: docs/versioning.md
• Release confidence gate: docs/release-confidence-gate.md
• Chinese workflow prompts: docs/prompts/zh-CN/ai-task-prompts.zh-CN.md
• Chinese user guide: docs/guides/zh-CN/user-guide.zh-CN.md
• Foundation implementation plan: docs/plans/implementation-plan.md

Repository Layout

Main source and test areas:

• src/: implementation source tree
• entrypoints/: WXT runtime entrypoints
• tests/: unit, integration, and e2e verification
• memory-bank/: canonical architecture memory
• docs/: repository-level guides and maintenance rules

Within docs/, Chinese prompts, guides, and plans are now grouped by purpose instead of being flattened at the repository root.

Maintenance Guardrails

• Keep modules split by runtime, domain, and feature.
• Do not collapse logic into monolithic files.
• Update memory-bank/@architecture.md after any architectural milestone or schema/message/runtime change.
• Prefer adding navigational and maintenance docs under docs/ instead of growing the root into a flat document pile.
• Follow docs/versioning.md for version bumps, commit expectations, and Edge build-output expectations.