v0.6.2 — Proportionate Agent Verification

Changed

• Made the narrowest sufficient verification the default, widening for required obligations, uncovered risks, or uncertain impact without weakening safety, integration, CI, release gates, or accountable human acceptance.
• Clarified evidence validity and reuse, conditional additional verification, and stopping once acceptance is supported instead of automatically duplicating proof.
• Bounded stalled work and recovery by outcomes rather than repeated activity, without treating exhausted effort as success.
• Added routine-verification routes from AGENTS.md, the main context map, and existing retrieval metadata; refined existing playbook, orchestration, and review guidance without a new canonical layer or checklist items.

Evaluation

• Reviewed long-form articles and vendor documentation, excluding standalone tweets; model-specific guidance, evaluation tooling, and raw artifacts remain outside the vendored core.
• Preserved the initial eight-run decision-only pilot and its disclosed global-prompt contamination; it did not demonstrate an efficiency gain.
• Ran six executable synthetic-consumer sessions with identical captured effective prompts/tools and explicit isolation: four completed tasks and two correctly blocked mandatory-integration handoffs, with no observed scope or acceptance regression.
• The candidate consulted the verification policy in 3/3 runs versus 0/3 at baseline. Candidate elapsed times were higher in these observations; no speedup or human-review saving is claimed.
• No application-wide suites or same-check/same-input reruns were executed. An additional pre-fix reproduction supplied useful evidence rather than redundant verification.

Validation

• Passed nix shell nixpkgs#yq nixpkgs#ripgrep nixpkgs#python3 -c ./tooling/run-checks.sh.
• Passed git diff --check.

Evidence and limits

• Source review
• Executable evaluation and limitations

Full changelog: v0.6.1...v0.6.2

v0.6.1 — Explicit Policy Acceptance

Changed

• Separated accountable acceptance of consequential recovery policies from factual authority, edge-mechanism execution, reconciliation ownership, and bounded-context ownership of local state and consequences.
• Required recovery records to name the policy acceptor and declared scope, or keep the gap unknown and the policy decision open.
• Added a concise architecture check against inferring policy acceptance from component ownership, authority access, or mechanism execution.

Evaluation

• Against candidate 202b1fb, B4 preserved T=2 and obtained R3=2 in 2/2 runs with every applicable R criterion at 2.
• H7 preserved T=2, obtained G3=2 in 2/2 runs, and invented no owner or authority; one G5=1 remained at the prior floor and outside the candidate gate.
• Blinded scoring used a separate model ID from the same provider family; backend immutability and provider-independence limits remain explicit.

Validation

• Passed nix shell nixpkgs#yq nixpkgs#ripgrep nixpkgs#python3 -c ./tooling/run-checks.sh.
• Passed git diff --check.

v0.6.0 — Living, Bounded and Verifiable Domain Context

Added

• Canonical definitions for domain model and model expression, with operational authority bounded by situated knowledge and accountable human validation.
• Knowledge-provenance statuses, missing-perspective prompts, linguistic clues, contrasting scenarios, and evidence-sized learning cycles for collaborative discovery.
• Explicit as-is, to-be, and transition views for context maps.
• Published Language guidance, provider-neutral probabilistic boundaries, interaction charters, mode pivots, and carry-only-conclusions handoffs.
• Unknown outcomes, stable business intention across retries, fact authority, reconciliation, and claim-shaped evidence for distributed workflows.
• A DDD Europe 2026 source review and reproducible v0.5.0/v0.6.0 behavioral evaluation.

Changed

• Clarified that DDD remains the core while agent interaction and verification guidance enable discovery, application, and protection of accepted domain models.
• Tightened workflow and failure guidance so timeout, retry exhaustion, and idempotency mechanisms do not invent business outcomes.
• Derived test scope from responsibility, boundary, risk, and needed confidence rather than a universal test taxonomy.

Evaluation

• Repeated B1-B6 with the exact v0.5.0 protocol: 12/12 runs completed without error or stderr.
• Specific behavior improved from 20/24 to 24/24; the applicable transversal rubric improved from 160/172 to 168/172, with no observed score regression.
• Outputs above 450 words decreased from 2/12 to 0/12; B1's residual model/expression ambiguity, ceiling effects, and single-model limits remain documented.

Validation

• Passed nix shell nixpkgs#yq nixpkgs#ripgrep nixpkgs#python3 -c ./tooling/run-checks.sh.
• Passed git diff --check.

v0.4.0

v0.4.0 — 2026-05-13

Added

• Agent-operability guidance for consumer repositories:
  • agent-ready repository playbook
  • agent work orchestration patterns
  • agent-operability review checklist
  • ADR 0006 for the agent-operability boundary
• design deviation vocabulary and review guidance for justified, scoped departures from canonical rules.
• Source reviews for agent harness/orchestration, tenets v0.6, and mattpocock/skills.
• Pi runtime smoke benchmark report for the release candidate.

Changed

• Expanded quality gates, architecture review, ADR writing, and deep-module guidance.
• Added feature/spec pre-implementation prompts to domain modeling.
• Updated maps and manifests for the new agent-operability canonical docs.
• Kept Spec-Kit, slash-command skills, runner protocols, issue trackers, and provider-specific tooling outside the core pack.

Validation

• Passed repository checks with nix shell nixpkgs#yq nixpkgs#ripgrep -c ./tooling/run-checks.sh.
• Passed git diff --check.
• Pi smoke benchmark completed 24/24 runs without runtime errors or timeouts.

v0.3.0

Current release after repository rename and core-pack cleanup.

Highlights:

• Canonical public repository: oneingan/design-context
• Default vendored path for consumer repos: docs/vendor/design-context/
• Core context pack and vendoring boundary documented in VENDORING.md
• Agent-readable design context framework for DDD, workflows, patterns, playbooks, and review checklists
• Lightweight validation includes ripgrep in the validation environment
• Prior repository history and previous release milestones remain visible (v0.1.0, v0.2.0)

Release commit: 3dc28d62fa328d0bee330779e8cb1bb66ccdac5c.