Home/pyrite/Epic: Migrate kleptocracy timeline to Pyrite-managed KB

Epic: Migrate kleptocracy timeline to Pyrite-managed KBbacklog_item

migrationcascadeepiccapturecascade
2 min read · Edit on Pyrite

Overview

Migrate the kleptocracy timeline project (capturecascade.org) from its current ad-hoc script-based management into a Pyrite-managed knowledge base. The timeline contains 4,400+ events spanning 1142-2026, with ~1,235 unique actors, custom validation scripts, actor alias normalization, source quality frameworks, and a React + Hugo viewer deployed to GitHub Pages.

Current Architecture

Repo: github.com/markramm/CaptureCascadeTimeline (mono-repo)

  • `timeline/data/events/` — 4,400+ markdown files with YAML frontmatter (source of truth)
  • `timeline/data/actor_aliases.json` — 880-line alias mapping (~250 canonical actors)
  • `timeline/scripts/` — custom Python validation, normalization, API generation, QA audit
  • `timeline/viewer/` — React SPA consuming static JSON API files
  • `hugo-site/` — Hugo static site reading markdown directly
  • `.github/workflows/ci-cd.yml` — validate → build React + Hugo → deploy GitHub Pages
  • `.claude/agents/` — research-executor and quality-improver agent configs

    • Target Architecture

    KB managed by Pyrite (Cascade plugin):

  • Pyrite handles: validation, QA, search, normalization, actor management, MCP access
  • Actor entries as first-class KB entries with aliases and backlinks
  • Research agents create events via Pyrite MCP/CLI instead of raw file writes
  • QA includes source URL validation and content verification

    • Viewer remains custom (same repo or separate):

  • CI/CD runs `pyrite cascade export` to generate static JSON API files
  • React viewer and Hugo site consume exported data (unchanged)
  • Deployed to capturecascade.org via GitHub Pages

    • Dependencies (Backlog Items)

    1. Support string-or-ref actor fields in Cascade events — schema change to accept both formats 2. Actor alias suggestion and fuzzy matching tool — port alias detection from timeline project 3. Actor extraction and migration tool — create actor entries from event string references 4. Backlink indexing for string-based actor references — index string actors for backlinks/graph 5. Cascade timeline static export for viewer consumption — replace custom generate.py 6. Source URL liveness checking — verify source URLs return 200 7. Source content verification via LLM — verify sources support entry claims 8. AI hallucination detection — web search for independent corroboration

    Migration Steps

    1. Add `kb.yaml` to kleptocracy timeline repo pointing Pyrite at `timeline/data/events/` 2. Verify Pyrite indexes all 4,400+ events correctly 3. Extract actors into KB entries using migration tool 4. Replace `generate.py` with `pyrite cascade export` in CI/CD 5. Replace custom validation scripts with `pyrite qa validate` + `pyrite ci` 6. Update .claude/agents to use Pyrite MCP for event creation 7. Run source URL validation on existing events 8. Update pre-commit hooks to use Pyrite validation

    Success Criteria

  • All 4,400+ events indexed and searchable in Pyrite
  • Actor entries with backlinks to all referencing events
  • CI/CD pipeline uses Pyrite commands instead of custom scripts
  • capturecascade.org viewer works identically (no visible changes)
  • Research agents create events via Pyrite MCP
  • QA includes URL validation and source verification
  • No data loss during migration

    • Linked from

    Actor alias suggestion and fuzzy matching tool Actor extraction and migration tool for Cascade timelines Support string-or-ref actor fields in Cascade events Source URL liveness checking for QA (Phase 1) Backlink indexing for string-based actor references Cascade timeline static export for viewer consumption AI hallucination detection for research KBs (Phase 3) Source content verification via LLM (Phase 2)

    Links to

    Support string-or-ref actor fields in Cascade events Actor alias suggestion and fuzzy matching tool Actor extraction and migration tool for Cascade timelines Backlink indexing for string-based actor references Cascade timeline static export for viewer consumption Source URL liveness checking for QA (Phase 1) Source content verification via LLM (Phase 2) AI hallucination detection for research KBs (Phase 3)