Skip to content

README.md

Latest commit

 

History

History
103 lines (77 loc) · 4.26 KB

README.md

File metadata and controls

103 lines (77 loc) · 4.26 KB
Raw

ClaimBound Documentation Index

Use this page to choose a reading path. You do not need every protocol file to try ClaimBound.

Reviewer Path

For first-time readers, program reviewers, journalists and external inspectors:

  1. Reviewer path (one page) — fastest honest baseline
  2. ClaimBound in 30 seconds
  3. Reviewer summary — problem, baseline vs planned work, differentiation from adjacent projects
  4. Common misreadings
  5. Current evidence tracks — plain-language map of all registry cards
  6. ESA 500-card heatmap — interactive coverage map and method
  7. Evidence card examples
  8. Case studies
  9. Planned work not shipped — roadmap vs current code
  10. Artifacts catalog — NYC TLC / CDC artifact-only records
  11. Public roadmap 2026
  12. European Dimension

Closing VERIFY mirrors? See Volunteer one-pager or External verification packs. Issues #85–#87 are open for independent external operators.

Quick validation:

uv sync --extra dev
uv run claimbound validate-all
uv run --extra dev python -m pytest -q

Operator Path

For people who want to run or challenge a card locally:

  1. Start without coding — Windows/macOS/Linux, no jq or bash
  2. Platform support
  3. ClaimBound in 5 minutes
  4. Getting started
  5. External verification packs — split VERIFY work
  6. External operator starter pack
  7. Result status protocol
  8. Evidence card protocol
  9. Independent rerun workflow
  10. Artifacts catalog — not registry cards

Rerun playbooks: runbooks/.

Developer Path

For contributors extending validators, runners or card tooling:

  1. Contributing guide
  2. Evidence card protocol
  3. Manual audit protocol
  4. AI operator protocol
  5. Software development workflow

Advanced (Optional)

These layers help manage related tracks and complex R&D families. They are not required to read cards, run demos or validate the registry.

Examples (Not First-Screen Evidence)

Optional methodology patterns kept out of the main card table:

Pre-Implementation Specs

Acceptance criteria for planned infrastructure. These are design documents, not completed implementations. See Planned work not shipped.

Deep Dives (Optional, Not First-Screen)

These pages support specific audiences or future direction. They are not part of the default reviewer or operator path.

Document Role
Project next steps Maintainer focus notes
Global evidence registry direction Future hosted-registry vision only
Reproduction guide (legacy) NASA/NOAA background; use rerun workflow + tier C packs
Flagship Grok workflow Worked public-AI example
Audience and value Audience-specific positioning
Claims Claim-boundary language reference