Version 0.4.3

agent-devkit 0.4.3

This release makes repository conventions, workflow state, review evidence,
and whole-repository simplicity audits explicit across the coding workflow.

Changes

• Carries exact cross-task rules into plans through ## Global Constraints.
• Maps approved edge cases to implementation tasks and concrete verification.
• Allows session-only Ruling R<n> decisions for reversible,
  behavior-equivalent implementation details.
• Requires review findings to cite path:line and reports unverifiable
  requirements as failing cannot verify spec gaps.
• Calibrates spec and plan self-review against behavior, scope, correctness,
  and execution rather than stylistic preferences.
• Adds recommendation-depth and adaptive edge-case coverage to brainstorming.
• Captures repository conventions from local evidence with approval,
  provenance, and scoped enforcement.
• Adds standalone lean-audit for report-only, evidence-backed simplicity
  findings.
• Adds EVALS.md scenarios 25, 26, and 27 for the workflow improvements.

Verification

• npm test passed with the plugin smoke checks.
• git diff --check passed.
• Scenarios 25, 26, and 27 passed in fresh sessions against disposable
  repositories.

Version 0.4.2

agent-devkit 0.4.2

This release makes workflow state and review evidence explicit across
brainstorming, planning, implementation, and verification.

Changes

• Carries exact cross-task rules into plans through ## Global Constraints.
• Maps approved edge cases to implementation tasks and concrete verification.
• Allows session-only Ruling R<n> decisions for reversible,
  behavior-equivalent implementation details.
• Requires review findings to cite path:line and reports unverifiable
  requirements as failing cannot verify spec gaps.
• Calibrates spec and plan self-review against behavior, scope, correctness,
  and execution rather than stylistic preferences.
• Adds recommendation-depth and adaptive edge-case coverage to brainstorming.
• Adds EVALS.md scenario 25 for workflow state and review evidence.

Verification

• npm test passed with the plugin smoke checks.
• git diff --check passed.
• Scenario 25 passed in fresh sessions against a disposable repository.

Version 0.4.1

agent-devkit 0.4.1

This release improves how brainstorm-feature clarifies ambiguous work before
design and implementation.

Changes

• Builds an internal decision tree from the intended user, problem, success
  criteria, scope, permissions, data lifecycle, interfaces, failure recovery,
  rollout, and verification needs.
• Asks one material question at a time, with known facts and a recommended
  answer plus its main trade-off.
• Follows vague, partial, or contradictory answers with a concrete
  clarification instead of silently choosing an interpretation.
• Avoids asking for repository facts or low-impact implementation details that
  the agent can discover or that do not affect the design.
• Adds EVALS.md scenario 24 for deep feature clarification.

Verification

• npm test passed with the plugin smoke checks.
• git diff --check passed.
• The updated skill was exercised against a disposable repository and produced
  a concrete recommendation before asking for confirmation.

Version 0.4.0

agent-devkit 0.4.0

agent-devkit is a lightweight, prompt-driven workflow toolkit for coding
agents. It gives agents a consistent way to understand a repository, plan and
implement changes, debug issues, verify work, and keep an LLM-facing codebase
wiki aligned with the source.

This release packages the same 12 reusable Markdown skills for multiple coding
agent harnesses while keeping the canonical skills portable and dependency-free.

What is included

• using-devkit routes each request to the appropriate workflow.
• setup-codebase creates a project-specific agent contract and documentation
  skeleton.
• setup-openez and read-codebase-context support evidence-based codebase
  exploration, with direct-source fallbacks when OpenEZ is unavailable.
• brainstorm-feature, plan-feature, and implement-task provide a staged
  feature delivery workflow.
• systematic-debugging guides root-cause investigation before implementation.
• review-and-verify provides a final evidence-based review and verification
  gate.
• document-wiki maintains source-grounded LLM documentation under
  docs/llm/.
• context-handoff preserves useful state when work is paused or a session is
  compacted.
• estimate-feature optionally produces per-task effort ranges when an
  estimate is requested.

Native harness packaging

The repository includes optional adapters for Codex, Claude Code, Cursor,
Devin CLI, and OpenCode. These adapters reuse the canonical skills/ tree and
provide harness-specific discovery and bootstrap behavior where supported.

The shared bootstrap loads using-devkit at session start for hook-based
harnesses. The OpenCode adapter registers the skills directory and injects the
bootstrap into the first user message. Skill references work in both
namespaced plugin installations and direct skills installations.

Design goals

• Keep workflow instructions readable, reviewable, and easy to customize at
  the source level.
• Prefer repository evidence over guesses when tracing code or documenting
  behavior.
• Keep changes small, explicit, and verifiable.
• Separate design, implementation, debugging, and review responsibilities.
• Avoid runtime dependencies and build tooling for the canonical Markdown
  skills.

Verification

Version 0.4.0 includes dependency-free smoke checks covering hook output,
OpenCode bootstrap behavior, duplicate protection, skill registration, and
release metadata consistency. The Codex plugin manifest and Claude Code
marketplace manifest also pass their available validators.

The skills can be used directly from an Agent Skills-compatible directory, or
distributed through the native packaging for the supported harnesses.