codex-pdf
betaAuthoritative, versioned PDF facts contract for Print with Synergy tools.
Contract-first output · Schema-validated JSON · CLI and parity workflows
Authoritative PDF facts,
not fragmented guesses.
codexPDF is Print with Synergy's authoritative, read-only PDF facts reference. Extract once, validate against published schemas, and let downstream tools consume stable document facts.
The contract layer your PDF stack can trust
codexPDF centralizes extraction into one authoritative facts engine so every downstream system can operate from the same versioned truth.
Contract-first output
CodexDocument is the stable root contract for extracted PDF facts, independent of consumer product logic.
Schema validation
Published schemas under versioned paths let every payload be validated in CI and runtime workflows.
Read-only boundary
codexPDF extracts facts only; no edits, no rendering mutations, no hidden policy side effects.
CLI workflows
Use extract, probe, validate, and parity commands to wire codex into local tooling and automation.
Parity profiles
Projection-based parity checks compare codex output against baseline systems and highlight drift.
Consumer-agnostic
Downstream tools can adapt codex output through thin adapters instead of re-parsing every document.
Typed Python models
Python package ships typed models and contract-aware primitives for reliable integrations.
Open-source AGPL
Full source under AGPL-3.0-or-later with no closed black-box extraction layer.
Open source
One family. Four repos. Choose your layer.
Build your stack from focused components: facts extraction, preflight checks, embeddable review, and document assay.
loupe-pdf
betaEmbeddable PDF viewer with separations, TAC, layers, and annotation overlays.
lint-pdf
betaDetection-only PDF preflight engine with deep standards coverage.
assay-pdf
betaPDF assay and metadata reporting for structural inventory.