Skip to content

Latest commit

 

History

History
46 lines (32 loc) · 3.32 KB

File metadata and controls

46 lines (32 loc) · 3.32 KB

Changelog

All notable changes to ContextFence are documented in this file.

The format is based on Keep a Changelog, and this project follows Semantic Versioning.

0.2.0 - 2026-07-18

Added

  • Declarative matrix block that expands an authorization model (identities, sources, per-source allow lists, and canaries) into the full identity × source grid of deterministic probes: a critical deny probe for every unauthorized identity and a positive control for every authorized one. Generated probes compile down to ordinary v1 probes and report against their source line. Includes examples/contracts/matrix.boundary.yaml and examples/contracts/matrix-leak.boundary.yaml.
  • contextfence generate command that turns a connector-neutral access manifest (identities, sources, and per-source allow lists) into a matrix boundary contract, deriving probe keys and canaries deterministically and emitting environment placeholders for target and identity credentials. Includes examples/manifests/northstar-access.yaml.
  • A clearer interactive lab path: one cache-isolation failure by default, scroll-to-result behavior, published npm and GitHub Action quick starts, and an explicit Boundary Baseline implementation offering.
  • CITATION.cff, a sitemap, and expanded social metadata for discovery and research use.

Changed

  • The package homepage now leads to the Devector ContextFence product page; funding links lead to the public assessment scope rather than a generic contact endpoint.

0.1.0 - 2026-07-11

Added

  • Interactive, responsive regression lab for exploring RAG permission failures and remediations.
  • Versioned YAML boundary contracts with identity, target, probe, severity, and deterministic assertion definitions.
  • contextfence test CLI with environment interpolation, dry runs, target overrides, concurrency, timeouts, severity thresholds, and explicit exit codes.
  • Mock and OpenAI-compatible target adapters.
  • Pretty, JSON, JUnit, SARIF, and portable HTML reports with redaction-aware evidence handling.
  • Source, citation, response, regex, and containment assertions over observable target evidence.
  • Reusable action.yml for running pinned ContextFence versions in GitHub Actions.
  • Synthetic example contracts, fixture documents, and a local mock target.
  • Production non-root container, multi-architecture GHCR publishing, and GitHub Pages demo deployment.
  • npm trusted-publishing workflow with registry provenance, GitHub artifact attestation, checksum, packed-CLI smoke test, and generated GitHub Release.
  • CI, CodeQL analysis, Dependabot updates, issue forms, pull request template, security policy, contribution guide, code of conduct, and release runbook.

Security

  • Target credentials are resolved from environment variables rather than committed contracts.
  • Machine-readable reports minimize request headers and support safe, synthetic reproduction workflows.
  • Release automation uses short-lived OIDC credentials instead of a long-lived npm publishing token after the one-time package bootstrap.