Files
Orchestrator/privat/CV/.kiro/specs/personal-knowledge-base/tasks.md
T
ankn a5f8fb49ab Migrate all repos into monorepo context folders
Bahn: aisupport, Analyse-O2C-C2S, awesome-bahn-mcp-servers, beam-mcp,
      Confluence_Bot, db-planet-mcp-server, O2C-Harness, project-audit,
      Projekt-KIQ-HP, teamlandkarte-mcp
Dhive: Jury-Voting
Privat: CV, NoteGraph (NOTE: NoteGraph needs complete redo after consolidation)
Shared: AI-Orchestrator, OrgMyLife, power_skills_and_more
Shared/references: symphony (read-only)

Bahn repos remain available as independent remotes - this monorepo
pulls them in via subtree, the originals are untouched.
2026-06-30 20:39:52 +02:00

5.7 KiB
Raw Blame History

Implementation Plan: Personal Knowledge Base

Overview

File-based personal knowledge graph with YAML entity storage, a graph index, and four Kiro skills. TypeScript for validation logic and property-based tests (fast-check). File system is the database — no build step or server required.

Source of truth: SPEC.md

Tasks

  • 1. Set up project structure and core configuration

    • 1.1 Create directory structure for the knowledge base
    • 1.2 Initialize TypeScript project for validation and testing
    • 1.3 Create README.md documenting the project
  • 2. Implement entity schemas, validation, and property tests

    • 2.1 Define TypeScript interfaces for all entity types

      • src/schemas/types.ts
      • Requirements: 1.1, 1.2, 2.1
    • 2.2 Implement entity validation functions

      • src/schemas/validate.ts
      • Requirements: 1.3, 1.4, 9.2
    • 2.3 Write schema property tests (Properties 1, 2, 3, 15)

      • Single test file: src/schemas/validate.test.ts
      • Property 1: Entity serialization round-trip
      • Property 2: Entity schema validation (accepts valid, rejects invalid)
      • Property 3: Temporal consistency (start ≤ end)
      • Property 15: File naming convention (kebab-case)
      • Requirements: 1.1, 1.3, 1.4, 2.1, 8.2, 9.2
  • 3. Implement YAML serialization and file I/O layer

    • 3.1 Create YAML serialization/deserialization utilities

      • src/io/yaml-utils.ts
      • Requirements: 8.2, 2.1
    • 3.2 Create entity file management functions

      • src/io/entity-files.ts
      • Requirements: 9.2, 1.5
  • 4. Implement graph index management and property tests

    • 4.1 Create graph index read/write logic

      • src/graph/index-manager.ts
      • Requirements: 1.2, 9.4
    • 4.2 Implement graph index rebuild from entity files

      • Scan entity directories, rebuild kb/graph-index.yaml, detect broken references
      • Requirements: 1.2, 9.4
    • 4.3 Write graph property tests (Properties 4, 5)

      • Single test file: src/graph/index-manager.test.ts
      • Property 4: Graph index consistency
      • Property 5: Tandem entity integrity
      • Requirements: 1.2, 2.3, 2.4, 9.4
  • 5. Checkpoint — run all tests, verify green

    • npm test
  • 6. Implement interview-collect Kiro skill

    • 6.1 Create interview-collect SKILL.md

      • .kiro/skills/interview-collect/SKILL.md
      • Requirements: 4.1, 4.2, 4.3, 4.4, 4.5, 9.3
    • 6.2 Implement gap detection and conflict detection logic

      • src/interview/gap-detection.ts
      • src/interview/conflict-detection.ts
      • Requirements: 4.4, 4.5
    • 6.3 Write interview property tests (Properties 6, 7, 8)

      • Single test file: src/interview/interview.test.ts
      • Property 6: Interview session question limit (≤ 10)
      • Property 7: Gap detection completeness
      • Property 8: Conflict detection
      • Requirements: 4.2, 4.4, 4.5
  • 7. Implement cv-generate Kiro skill

    • 7.1 Create cv-generate SKILL.md and formatting reference

      • .kiro/skills/cv-generate/SKILL.md
      • .kiro/skills/cv-generate/references/cv-formatting-guide.md
      • Requirements: 5.15.6, 9.3
    • 7.2 Implement relevance selection and tandem generation logic

      • src/cv/relevance-selection.ts
      • src/cv/tandem-generator.ts
      • Requirements: 5.1, 5.2, 5.3, 5.4
    • 7.3 Create CV templates

      • templates/cv-individual.md
      • templates/cv-tandem.md
      • Requirements: 5.5
    • 7.4 Write CV property tests (Properties 9, 10, 11)

      • Single test file: src/cv/cv.test.ts
      • Property 9: CV relevance selection ordering
      • Property 10: Tandem application document set (exactly 3 docs)
      • Property 11: CV output validity and traceability
      • Requirements: 5.15.6
  • 8. Implement talk-intro Kiro skill

    • 8.1 Create talk-intro SKILL.md and template

      • .kiro/skills/talk-intro/SKILL.md
      • templates/intro-speaker.md
      • Requirements: 6.1, 6.2, 6.3, 9.3
    • 8.2 Implement introduction generation logic

      • src/intro/intro-generator.ts
      • Requirements: 6.1, 6.2, 6.3
    • 8.3 Write intro property test (Property 12)

      • Single test file: src/intro/intro.test.ts
      • Property 12: Introduction length variants (short < medium < long)
      • Requirements: 6.2
  • 9. Implement kb-review Kiro skill

    • 9.1 Create kb-review SKILL.md

      • .kiro/skills/kb-review/SKILL.md
      • Requirements: 7.1, 7.2, 7.3, 7.4, 9.3
    • 9.2 Implement completeness scoring and prioritization logic

      • src/review/completeness-score.ts
      • src/review/prioritization.ts
      • Requirements: 7.2, 7.3
    • 9.3 Write review property tests (Properties 13, 14)

      • Single test file: src/review/review.test.ts
      • Property 13: Completeness score calculation
      • Property 14: Review prioritization ordering
      • Requirements: 7.2, 7.3
  • 10. Checkpoint — run all tests, verify green

    • npm test
  • 11. Wire everything together

    • 11.1 Create KB conventions steering file

      • .kiro/steering/kb-conventions.md
      • Requirements: 9.2, 9.4
    • 11.2 Create seed data (graph-index.yaml + profile stub)

      • kb/graph-index.yaml (empty)
      • kb/profiles/andre-knie.yaml (minimal valid profile)
      • Requirements: 1.5, 9.1
    • 11.3 Wire skills to reference shared validation logic

      • Requirements: 9.3
  • 12. Final checkpoint — run all tests, verify green

    • npm test

Notes

  • Property tests consolidated into per-module test files (fewer files, same coverage)
  • SPEC.md is the source of truth; README.md is user-facing documentation
  • Seed data (11.2) runs after validation logic exists so it can be validated
  • AGENTS.md git workflow (branch, PR) applies at commit time