Changelog

All notable changes to Agentic InfraOps are documented in this file.

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

[0.10.0] — Unreleased

Added

feat(diagrams): add Excalidraw-native Azure icon libraries, reference docs, icon conversion tooling, and Excalidraw validation/export automation for architecture artifacts.
feat(ci): add Excalidraw SVG export and Azure icon update workflows, plus devcontainer/editor support for Excalidraw-based diagram authoring.
feat(skills): add workflow-engine skill with machine-readable DAG (workflow-graph.json) for graph-based step routing in the Conductor, replacing hardcoded step logic.
feat(skills): add context-shredding skill with 3-tier runtime compression (full/summarized/minimal) and per-artifact compression templates.
feat(session-resume): upgrade session state schema from v1.0 → v2.0 with atomic claim-based lock model (lock.owner_id, lock.heartbeat, lock.attempt_token), per-step claim objects, stale_threshold_ms, and event_log audit trail. Backwards-compatible with v1.0 files.
feat(skills): add circuit breaker pattern to iac-common skill with failure taxonomy (6 categories), anomaly detection thresholds, and mandatory stopping rules for deploy agents.
feat(skills): add Smart PR Flow to github-operations skill with label-based lifecycle tracking, auto-label rules, and deploy agent watchdog integration.
feat(config): add .github/agent-registry.json — machine-readable registry mapping agent roles to definition files, default models, and required skills.
feat(config): add .github/skill-affinity.json — skill/agent affinity catalog with primary/secondary/never weights for context budget optimization.
feat(scripts): add 5 new validators: validate-session-lock, validate-workflow-graph, validate-agent-registry, validate-skill-affinity; validator count 22 → 25.
feat(hooks): add pre-push hook to lefthook.yml with diff-based domain routing; only runs validators for changed file types, in parallel.
feat(scripts): add diff-based-push-check.sh helper for pre-push hook domain detection.
feat(skills): add terraform-test skill with run blocks, mock providers (TF 1.7+), assertions, CI/CD patterns, and Azure examples (azurerm_resource_group, azurerm_virtual_network, azurerm_key_vault).
feat(skills): add terraform-search-import skill with manual discovery workflow, 20-row ARM↔Terraform mapping table, and bulk import script.
feat(skills): extend terraform-patterns with 2 new references: avm-authoring-requirements.md (37 AVM certification checks) and refactor-module.md (monolith-to-module extraction with state migration).
feat(skills): add count-registry skill for agent runtime entity count lookups from count-manifest.json.
feat(config): add .github/count-manifest.json as single source of truth for entity counts (agents, skills, instructions, validators).
feat(instructions): add no-hardcoded-counts.instructions.md to prevent hard-coded entity counts across all markdown, JSON, and script files.
feat(scripts): add validate-no-hardcoded-counts.mjs validator.
feat(agents): add RALPH-style E2E autonomous workflow conductor (e2e-conductor.agent.md) with pre-validation after every subagent return, auto-challenge (1 pass per step), self-correction, benchmark collection, and lesson capture — no human gates.
feat(e2e): add Terraform support to Ralph Loop E2E testing — prompt parameterization with IaC tool routing, e2e-conductor dual-track (Bicep + Terraform subagents), Terraform validators in validate-e2e-step.mjs (terraform validate, terraform fmt -check), and Terraform scoring in benchmark-e2e.mjs (scoreCodeQuality() with AVM-TF detection).
feat(e2e): add e2e:validate and e2e:benchmark npm scripts for E2E testing.
feat(e2e): register e2e-conductor in agent-registry.json.
feat(e2e): add e2e-validation.yml CI workflow (manual dispatch + weekly schedule) with benchmark report upload. Add structural E2E validation step to lint.yml.
feat(e2e): add configurable benchmark threshold via E2E_PASS_THRESHOLD env var.
feat(e2e): add multi-project benchmark comparison (--compare flag) with auto-discovery.
docs(e2e): add dedicated E2E testing documentation (docs/e2e-testing.md) with IaC track matrix, benchmark dimensions, and troubleshooting guide.
docs(e2e): add Ralph Loop, E2E Benchmark, E2E Conductor to glossary.
docs(e2e): add E2E testing guidance to AGENTS.md and CONTRIBUTING.md.
feat(e2e): add E2E evaluation scripts: benchmark-e2e.mjs (8-dimension scoring engine with complexity-normalized baselines) and validate-e2e-step.mjs (per-step validator orchestrator composing existing lint/validate commands).
feat(e2e): add E2E prompts: e2e-ralph-loop.prompt.md (RALPH-style 8-phase loop driver), e2e-analyze-lessons.prompt.md (post-loop lessons analysis), and e2e-contoso-rfp.prompt.md (full 7-step loop for Contoso Service Hub).
feat(e2e): complete e2e-ralph-loop (Nordic Fresh Foods Lite) end-to-end — all 7 steps with design diagrams, ADR, governance constraints, deployment summary, as-built documentation suite (7 docs), Bicep templates (main + 6 modules), benchmark score 88/100, and lessons-learned artifacts.
feat(e2e): bootstrap Contoso Service Hub RFP loop (contoso-service-hub-run-1) with session state, handoff, and benchmark scaffolding.
feat(docs): add interactive D3 architecture explorer (docs/architecture-explorer.html) — self-contained HTML visualization of the full system topology (agents, subagents, skills, instructions, validators, workflow gates, MCP servers, prompts, CI workflows) with force-directed layout, search, filtering, and view presets.

Changed

refactor(diagrams): switch the default architecture diagram workflow from Draw.io to Excalidraw across skills, prompts, agent instructions, VS Code MCP configuration, templates, and docs.
refactor(docs): update demo content and documentation to publish .excalidraw artifacts instead of .drawio files.
refactor(conductor): replace hardcoded step table with graph-based routing via workflow-graph.json and agent registry lookups.
refactor(conductor): add circuit breaker principle — halt on blocked step status.
refactor(agents): add context-shredding skill reference to Architect, Bicep CodeGen, Terraform CodeGen, and As-Built agents for runtime context compression.
refactor(agents): add iac-common/references/circuit-breaker.md and Smart PR Flow references to Bicep Deploy and Terraform Deploy agents.
refactor(instructions): add runtime compression and skill affinity sections to context-optimization.instructions.md.
refactor(agents,skills): agent system consistency pass across all agent files — standardize model field to array format, rename 8-step workflow references to multi-step, add Fast Path and Governance entries to skill-affinity.json, add session-resume to design/deploy agents in agent-registry.json.
refactor(agents): enhance 10-Challenger with multi-pass routing via challenger-review-batch-subagent, lens rotation table (security-governance, architecture-reliability, cost-feasibility), and pass-number routing (single → subagent, multi → batch).
refactor(agents): Conductor gate naming (Gates 1–5 + Gate 2.5 Governance) and Switch to Fast Path handoff for simple projects.
refactor(skills): context optimization of 7 skills — extract verbose content to reference files, reducing SKILL.md sizes by 40–70%: azure-cost-optimization, azure-quotas, context-optimizer, github-operations, azure-adr, azure-kusto, make-skill-template.
refactor(agents,scripts,instructions): consolidate instructions (#255) — merge bicep/terraform policy compliance into iac-policy-compliance.instructions.md, merge code-commenting + code-review into code-quality.instructions.md, move workload-documentation and cost-estimate to skill references.
refactor(scripts): merge 6 validators into 3: agent-body-size + agent-language → lint-agent-checks, skill-size + skill-digests → validate-skill-checks, instruction-frontmatter + instruction-references → validate-instruction-checks.
refactor(scripts): extract shared _lib/h2-parser.mjs utility; add getBody() export to parse-frontmatter.mjs; reformat workspace-index.mjs for readability.
refactor(skills): merge terraform-patterns style-guide delta into tf-best-practices-examples.md (code formatting, version control, code review sections).
refactor(e2e): parameterize E2E scripts for multi-project support — benchmark-e2e.mjs accepts project name via argv, validate-e2e-step.mjs accepts --project=name flag.
refactor(docs,scripts): update D3 explorer and reformat digest generator.
docs: replace 105 hard-coded entity counts across 45+ files with descriptive language; resolve ‘7-step’ vs ‘8-step’ workflow conflict (all refs now say ‘multi-step’).

Removed

chore(diagrams): remove the Draw.io MCP server, Draw.io icon libraries, Draw.io export scripts, and Draw.io demo assets in favor of the Excalidraw pipeline.

Fixed

fix(session-state): accept both schema_version "1.0" and "2.0" in validator for backwards compatibility.
fix(skills): remove 19 duplicate nested skill directories introduced by Azure skills plugin integration — each had a redundant self-named subdirectory shadowing the canonical content at the parent level.
fix(agents,skills): resolve tech debt #10, #15, #16 — convert multi-line agents frontmatter to inline arrays (8 agents), add Reference Index sections to 19 skills, add canary markers to 76 reference files.
fix(agents,skills): resolve 26 artifact template drift warnings — add collapsible ToC, traffic-light indicators, Mermaid diagrams, details blocks; fix H2→H3 demotions; add attribution header to deployment-summary template.
fix(prompts): correct frontmatter key from mode to agent in e2e-ralph-loop prompt.
fix(skills): fix trailing spaces and blank lines in SKILL.minimal.md and SKILL.digest.md files.

[0.9.0.1] — 2026-03-15

Added

feat(skills): integrate 22 skills from Azure Skills Plugin (microsoft/azure-skills): appinsights-instrumentation, azure-ai, azure-aigateway, azure-cloud-migrate, azure-compliance, azure-compute, azure-cost-optimization, azure-deploy, azure-diagnostics, azure-hosted-copilot-sdk, azure-kusto, azure-messaging, azure-prepare, azure-quotas, azure-rbac, azure-resource-lookup, azure-resource-visualizer, azure-storage, azure-validate, copilot-customization, entra-app-registration, microsoft-foundry.
feat(skills): add SKILL.digest.md and SKILL.minimal.md variants for all new skills.
feat(agents): implement cross-agent decision logging (#250) — add decision_log[] field to session state schema, decision-logging.instructions.md, and propagate to 6 agents (Requirements, Architect, Bicep/Terraform Planners, Bicep/Terraform CodeGen) plus challenger-review-subagent.
feat(instructions): add model-prompt-alignment.instructions.md — auto-applies to *.agent.md and *.prompt.md with model-specific prompt engineering patterns for Claude (selective XML blocks, reasoning_effort, language calibration) and GPT (structured markdown, tool-call-first phrasing), plus cross-model rules for handoff overrides, prompt-agent model sync, and few-shot guidance.
feat(scripts): add lint-model-alignment.mjs validator with 5 checks: prompt↔agent model sync, redundant handoff model overrides, Claude reasoning_effort comments, large-agent context_awareness, and investigate block presence; registered in validate:_node suite and lefthook.yml pre-commit hook.
feat(scripts): add generate-skill-digests.mjs for automated digest generation.
feat(docs): add Azure Skills Plugin migration documentation.

Changed

refactor(agents): align 8 Claude Opus/Sonnet agent definitions with Anthropic prompting best practices — add selective XML blocks (investigate_before_answering to 5 agents, output_contract to 5, context_awareness to 3, subagent_budget to Conductor, scope_fencing to 3, empty_result_recovery to Diagnose + 2 subagents), add reasoning_effort comments to 8 agents, add few-shot examples to Conductor/Architect/Planners.
refactor(agents): align 6 GPT-5.4/5.3-Codex agent definitions with OpenAI prompting best practices — add structured <output_contract> blocks, tool-call-first phrasing, and explicit phase-numbered workflows.
refactor(agents): conservative language softening across 8 agents — reduce duplicate absolute language (MANDATORY, NEVER, CRITICAL) by ~30% while preserving constraint emphasis at security baseline, approval gates, and governance compliance.
refactor(prompts): enhance 14 prompt files — fix 5 model mismatches (Claude/Sonnet→GPT-5.4), add prerequisites/variables/session-state-detection to 9 Claude prompt files.
refactor(agents): update 06b-Bicep CodeGen, 07b-Bicep Deploy agent definitions with enhanced deployment patterns and skill references.
refactor(config): update agent-registry.json, skill-affinity.json, and copilot-instructions.md for new skill integrations.

Fixed

fix(agents): resolve 19 handoff inconsistencies identified by dual adversarial review (Claude Sonnet 4.6 + GPT-5.4 reviewers):
- Critical: fix 09-Diagnose model mismatch (registry Sonnet→Opus to match frontmatter), remove wrong model override on 04-Design→Governance handoff (Sonnet→removed, target is GPT-5.4), add missing “Return to Step 2” handoff to 07t-Terraform Deploy (symmetry with 07b-Bicep Deploy).
- High: gate 04-Design skip paths with send: false and risk warnings (was bypassing mandatory governance + planning), redirect 05b/05t “Refresh Governance” to 04g-Governance (was self-handoff bypassing governance agent), add challenger-review-subagent to 04g-Governance agents list (was claiming review but couldn’t execute it), add 5 missing step handoffs to Fast Path conductor, add 10-Challenger to 02-Requirements agents list, change interactive Requirements handoffs to send: false.
- Cleanup: wire orphaned bicep-whatif-subagent and terraform-plan-subagent to their deploy agents, redirect Diagnose workload-docs handoff to 08-As-Built, fix governance planner prompts to reference both .md and .json artifacts, remove 05t-Terraform Planner from 03-Architect agents list, remove 5 redundant model overrides across 04-Design/04g-Governance/07b-Deploy.
fix(agents): remove 12 stale handoff model overrides from Conductor (9), Architect (2), and Requirements/Diagnose/Planners (3) — overrides were either redundant (matching target model) or pointing to wrong models after prior refactoring.
fix(hooks): fix VS Code hook scripts — correct API field names (toolName→tool_name, toolInput→tool_input), update tool name patterns to match actual VS Code tool IDs, use permissionDecision:deny for dangerous command blocking.
fix(prompts): fix diagnose-resource.prompt.md model field (Sonnet→Opus to match agent).

Removed

chore(skills): remove deprecated azure-diagnostics legacy skill (replaced by azure-diagnostics from Azure Skills Plugin).

Removed

chore(mcp): remove microsoft-learn MCP server from .vscode/mcp.json and .devcontainer/post-create.sh — Learn MCP tools are now bundled with the ms-azuretools.vscode-azure-github-copilot extension.
chore(skills): delete microsoft-docs, microsoft-code-reference, and microsoft-skill-creator skills — functionality provided natively by extension.
chore(config): remove skill references from agent-registry.json and skill-affinity.json.

⬆️ Back to Top## [0.9.0] — Pre-Bosun Baseline

Added

feat(terraform): add end-to-end Terraform IaC workflow with Planner (05t), Code Generator (06t), Deploy (07t), and supporting lint/review/plan subagents.
feat(agents): add IaC-track selection in Requirements and Terraform routing in Conductor.
feat(quality): add Terraform quality gates, CI checks, and IaC-neutral artifact template support.
feat(instructions): add Terraform best-practices and policy-compliance instruction coverage.
feat(skills): add terraform-patterns skill for AVM-TF composition and common pitfalls.
feat(conductor): add 00-handoff.md phase handoff/resume flow and related guidance.
feat(agents): add compact_for_parent summary output for challenger review passes.
feat(devcontainer): add Terraform toolchain and post-start updates for checkov, ruff, and diagrams.
feat(mcp): add GitHub and Microsoft Learn remote MCP servers.
feat(pricing-mcp): deliver Azure Pricing MCP v4.0/v4.1.0 improvements.
feat(scripts): add 5 CI enforcement validators for context-optimization guardrails (lint:agent-body-size, lint:glob-audit, lint:skill-size, lint:skill-references, lint:orphaned-content); validator count 15 → 22.
feat(agents): add fast-path 01-Conductor (Fast Path) agent for simple 1–3 resource projects with combined Plan+Code step and single challenger pass.
feat(ci): add weekly doc-freshness cron workflow and quarterly context audit checklist in AGENTS.md.

Changed

refactor(agents): simplify challenger review context handoff by switching parent retention to compact findings summaries.
chore(agents): rename/renumber agent files and align subagent orchestration patterns.
refactor(tf-dev): iterate Terraform phase prompts, routing, and subagent guidance.
docs: refresh README, prompt docs, and agent metadata for Terraform-first multi-track workflow.
chore(docs): remove legacy Terraform planning docs (docs/tf-support/, roadmap) after integration.
refactor(azure-mcp): migrate extension references to vscode-azure-mcp-server.
ci: tighten policy/compliance validation and branch merge-gate enforcement for Terraform rollout.
style: apply broad formatting and consistency cleanup across docs/instructions/scripts.
refactor(skills): split 10 large skills into core SKILL.md + on-demand references/ files; 60 reference files total; skill context load reduced by 46% vs M1 baseline.
refactor(agents): trim all agent bodies to ≤350 lines with explicit tool-boundary declarations; deduplicate cross-agent content via shared instruction globs; agent context reduced by 18%.
refactor(instructions): split 5 large instruction files into references/ sub-documents and enforce narrow glob patterns; instruction context reduced by 32%.
refactor(agents): overhaul subagent delegation patterns and introduce iac-common skill to consolidate shared Bicep/Terraform deploy logic across agents 07b and 07t.
feat(agents): upgrade challenger-review-subagent from GPT-4o to Claude Sonnet 4.6.
fix(frontmatter): convert YAML block scalar descriptions to single-line inline strings across all agent and skill frontmatter.

Fixed

fix(devcontainer): correct Terraform MCP server runtime path and move from Docker-based startup to Go binary execution.
fix(mcp): suppress Terraform Enterprise token noise on startup and correct MCP healthcheck behavior.
fix(agents): repair challenger/agent tool declarations and enforce subagent delegation rules.
fix(actions): move deprecation tracker automation to PR flow.
fix(validation): resolve validation drift for stable npm run validate:all execution.
fix(azure-pricing-mcp): align bulk estimate formatter with the indices response shape.
fix(scripts): add remediation messages to all context-optimization validator failure outputs.
fix(scripts): prevent YAML block scalar descriptions from recurring after frontmatter cleanup.
fix(agents): remove deprecated agent/runSubagent tool declarations from all 13 agents.
build(devcontainer): replace tfsec with checkov; pin tflint to v0.61.0.

⬆️ Back to Top## [0.9.0] - 2026-02-12

Changed

chore(version): reset project version from 0.0.45 to 0.9.0 for pre-production semantics.
chore(version): align VERSION.md, package.json, and pyproject.toml.
docs(version): make VERSION.md the documentation source of truth for version display.
build(version): simplify validate-version-sync.mjs checks to version-bearing files.

Note: Versions below (0.0.45 and earlier) are pre-release development milestones.

⬆️ Back to Top## [0.0.45] - 2026-02-05

Added

docs(agents): document model selection guidance in agent-definitions.instructions.md.

Changed

docs(readme): restructure README with collapsible sections and accessibility fixes.
docs(diagram): correct workflow sequence to show all five approval gates.

Fixed

fix(lint-yml): simplify markdown-lint trigger paths.
fix(skills): correct template paths in deployment preflight skill docs.
fix(templates): restore missing header text in governance constraints template.

⬆️ Back to Top## [0.0.44] - 2026-02-04

Added

docs(styling): add callouts and references sections across documentation.
chore(devcontainer-extensions): add mutantdino.resourcemonitor.
docs(terraform): add docs/terraform-roadmap.md and tracking issue #85.

Changed

ci(link-check): move to nightly schedule with issue auto-creation on failures.
chore(versioning): simplify manual release flow and remove auto-version workflow.

Fixed

docs(links): repair broad internal-link drift across docs and artifacts.
ci(workflows): fix workflow_dispatch input typing and increase link-check timeout.

⬆️ Back to Top## [0.0.43] - 2026-02-03

Added

feat(validation): add 9-category validation framework and validate:all workflow.
feat(skills): complete agent-to-skill migration with new ADR/workload docs capabilities.
ci(links): add external link checker workflow using lychee.

Changed

refactor(agents): reduce agent count by converting diagram, adr, and docs to skills.
docs(counts): remove hardcoded agent/skill totals in documentation.

Breaking Changes

chore(agents): remove @diagram, @adr, and @docs; replace with skill-based workflows.

⬆️ Back to Top## [0.0.42] - 2026-02-03

Changed

refactor(diagrams): standardize on Python diagrams library generation only.
build(diagrams): add diagrams, matplotlib, and pillow requirements.

Removed

chore(drawio): remove Draw.io MCP server, templates, scripts, and extension integration.

⬆️ Back to Top## [0.0.41] - 2026-02-02

Added

docs(plan): add agent-to-skill migration plan (plan-agentToSkillMigration.prompt.md).

Changed

docs(readme): overhaul README layout, navigation, and badge presentation.

Fixed

fix(skills): resolve markdown lint issues in deployment preflight and skill template files.

⬆️ Back to Top## [0.0.40] - 2026-01-23

Changed

feat(workflow): implement automated versioning and branch protection (#40).

⬆️ Back to Top## [0.0.39] - 2026-01-22

Added

feat(agent-testing): introduce complete agent validation framework.

⬆️ Back to Top## [0.0.38] - 2026-01-22

Changed

feat(agents): rename @plan to @requirements to avoid collision with VS Code built-in Plan.
refactor(agents): rename plan.agent.md to requirements.agent.md and update references.

⬆️ Back to Top## [0.0.37] - 2026-01-21

Added

feat(testing): add comprehensive agent testing plan prompt.

⬆️ Back to Top## [0.0.36] - 2026-01-21

Fixed

fix(devcontainer): resolve post-create permission issues.

⬆️ Back to Top## [0.0.35] - 2026-01-21

Fixed

fix(devcontainer): remove invalid PATH override that blocked container startup.

⬆️ Back to Top## [0.0.34] - 2026-01-21

Fixed

fix(docs): update remaining legacy agent references in embedded docs.

⬆️ Back to Top## [0.0.33] - 2026-01-21

Breaking Changes

feat(agents)!: rename agents to shorter verb-based names.

⬆️ Back to Top## [0.0.32] - 2026-01-21

Added

feat(agents): integrate deploy agent into the workflow.

⬆️ Back to Top## [0.0.31] - 2026-01-21

Breaking Changes

chore(terraform): remove Terraform support and move repository to Bicep-only operation.

Added

feat(git-hooks): replace Husky with lefthook.
docs(terraform): add Terraform re-enable guide at docs/guides/terraform-extension-guide.md.
chore(devcontainer-python): enable basic Pylance type checking.

Changed

fix(markdownlint): improve markdownlint detection in post-create checks.
chore(config): consolidate markdownlint config to .markdownlint-cli2.jsonc.
fix(mcp): replace unreliable stdio healthcheck with Python import verification.
docs(terraform): update repository docs to remove Terraform assumptions.

Removed

chore(husky): remove Husky directory and dependency.
chore(terraform): remove Terraform tooling, references, and related config entries.

⬆️ Back to Top## [0.0.30] - 2026-01-20

Added

feat(diagnose): add Azure Resource Health Diagnostician agent.

⬆️ Back to Top## [0.0.29] - 2026-01-19

Fixed

fix(ci): correct version auto-update extraction logic.

⬆️ Back to Top## [0.0.28] - 2026-01-19

Added

feat(deploy-agent): activate deploy agent for Step 6 workflows.
docs(deploy-agent): add dual-path deployment guidance and troubleshooting notes.

Changed

docs(workflow): update diagrams and references to use Deploy Agent terminology.

⬆️ Back to Top## [0.0.27] - 2026-01-19

Added

feat(presenter): add dark-themed workflow diagram for presentations.

⬆️ Back to Top## [0.0.26] - 2026-01-19

Changed

chore(release): prepare release transition to 0.0.25 baseline.

⬆️ Back to Top## [0.0.25] - 2026-01-19

Breaking Changes

refactor(scenarios): renumber and reduce scenarios from 11 to 8.

Changed

docs(workflow): consolidate workflow docs to docs/reference/workflow.md.
docs(cleanup): remove duplicate guides and standardize budget terminology.
chore(paths): update scenario references and paths across the repo.

Removed

chore(legacy): remove scenarios/scenario-output/ and legacy docs folders.
chore(example): remove infra/bicep/contoso-patient-portal/.

⬆️ Back to Top## [0.0.24] - 2026-01-14

Added

feat(demo): add prompt for 30-minute live workflow demo.

⬆️ Back to Top## [0.0.23] - 2026-01-14

Fixed

fix(prompts): convert plan-requirements to proper prompt-file format.

⬆️ Back to Top## [0.0.22] - 2026-01-14

Added

feat(artifacts): complete artifact template compliance rollout.

⬆️ Back to Top## [0.0.21] - 2026-01-14

Added

feat(artifacts): reach standard strictness across all 12 artifact types.

Changed

refactor(wave2): align all 07-* artifacts with template structure.
refactor(legacy): align ecommerce legacy artifacts with current templates.
chore(strictness): raise validation strictness from relaxed to standard.
chore(validation): expand allowed optional sections for common additions.

Fixed

fix(package): remove duplicate version line.
fix(docs): remove outdated design document TOC sections.

⬆️ Back to Top## [0.0.20] - 2026-01-14

Added

feat(templates): add 8 new artifact templates for governance, implementation, and as-built outputs.
feat(validation): add per-artifact strictness configuration.

Changed

refactor(validation): generalize Wave 1 validation to all 12 artifact types.
docs(readme): redesign workflow tables and legend.
chore(artifacts): rename ecommerce artifacts to standard naming convention.
ci(workflows): expand trigger paths for templates and agent changes.

Fixed

fix(docs): correct renamed artifact references in ecommerce documentation index.

⬆️ Back to Top## [0.0.19] - 2026-01-14

Added

feat(deploy): add Step 6 deploy agent.
feat(skills): add GitHub issues skill and template drift guards.
feat(docs): introduce docs/reference/, docs/getting-started/, and merged presenter docs.
feat(validation): add Wave 1 artifact and cost estimate template validation pipelines.

Changed

refactor(project-planner): align planner workflow/tooling with modern custom agent patterns.
chore(agents): standardize shared defaults and relative template links across agents.
chore(validation): increase Wave 1 strictness to standard.

Fixed

fix(tools): update deprecated tool-name references in agent docs.
fix(links): resolve markdown lint and broken-link issues across instruction and artifact files.

Removed

chore(terraform-doc): remove obsolete terraform-azure.instructions.md.
chore(docs): merge and remove docs/presenter-toolkit/ and docs/value-proposition/.

⬆️ Back to Top## [0.0.18] - 2026-01-13

Changed

refactor(agents): rename @plan display references to Project Planner across docs.
docs(usage): correct invocation guidance and regenerate workflow diagrams.

⬆️ Back to Top## [0.0.17] - 2025-12-18

Fixed

fix(pricing): update Azure Pricing Calculator URLs with locale-aware links.

⬆️ Back to Top## [0.0.16] - 2025-12-18

Fixed

fix(paths): correct relative paths in azure-principal-architect.agent.md.

⬆️ Back to Top## [0.0.15] - 2025-12-18

Fixed

fix(paths): correct shared foundation link path in all agents.

⬆️ Back to Top## [0.0.14] - 2025-12-18

Fixed

fix(readme): correct table link paths.

⬆️ Back to Top## [0.0.13] - 2025-12-18

Fixed

fix(readme): remove non-functional Mermaid click links and add link table.

⬆️ Back to Top## [0.0.12] - 2025-12-18

Fixed

fix(readme): switch Mermaid click links to absolute GitHub URLs.

⬆️ Back to Top## [0.0.11] - 2025-12-18

Fixed

fix(readme): correct Mermaid click links.

⬆️ Back to Top## [0.0.10] - 2025-12-18

Fixed

fix(docs): clean up docs rebuild path/link breakage.

⬆️ Back to Top## [0.0.9] - 2025-12-17

Added

feat(validation): add static-webapp-test workflow validation example.

⬆️ Back to Top## [0.0.8] - 2025-12-17

Changed

feat(workflow): integrate requirements template into the workflow.
refactor(workflow): restructure to 7-step lifecycle with Deploy as Step 6.
chore(artifacts): standardize -des and -ab artifact suffixes.
refactor(costing): move cost estimates to Step 3 design artifacts.
docs(pricing): add Azure Pricing MCP fallback chain guidance.

⬆️ Back to Top## [0.0.7] - 2025-12-17

Added

feat(diagrams): add workflow diagram generator setup.

⬆️ Back to Top## [0.0.6] - 2025-12-17

Added

feat(docs-agent): add workload documentation generator agent for optional Step 7.

⬆️ Back to Top## [0.0.5] - 2025-12-17

Added

feat(outputs): centralize agent outputs and automate versioning.

⬆️ Back to Top## [0.0.4] - 2025-12-07

Added

feat(personas): add character reference card for all personas.
ci(actions): add GitHub Actions workflow with five validation jobs.
feat(shared): add _shared/defaults.md configuration.
feat(scenarios): add healthcare, analytics, and static website demo scenarios.
docs(adr): add ADR-001 through ADR-004.
docs(roadmap): add project improvements plan.

Changed

refactor(scenarios): renumber scenarios S01-S11.
refactor(personas): resolve character naming collisions.
feat(pricing-mcp): improve caching, timeouts, and session handling.

Fixed

fix(scenarios): remove duplicate S04 folders.
fix(personas): fix character-name collisions across scenarios.
fix(mcp): resolve “Connector is closed” server errors.

⬆️ Back to Top## [0.0.3] - 2025-12-03

Changed

refactor(docs): reorganize docs into workflow, getting-started, guides, value-proposition, and cost-estimates subfolders.
refactor(scenarios): reorganize scenarios with quick-demos subfolder.

Breaking Changes (File Paths)

chore(paths): move docs/WORKFLOW.md to docs/workflow/WORKFLOW.md.
chore(paths): move docs/QUICKSTART.md to docs/getting-started/QUICKSTART.md.
chore(paths): move docs/troubleshooting.md to docs/guides/troubleshooting.md.

⬆️ Back to Top## [0.0.2] - 2025-12-01

Changed

refactor(repo): restructure repository around the 7-step agent workflow.
chore(structure): simplify folder layout by removing legacy scenario structure.

Added

feat(agents): add custom agents for Azure infrastructure workflow.
docs(workflow): add comprehensive workflow documentation.
feat(prompts): add e-commerce scenario prompts.
feat(pricing-mcp): add Azure Pricing MCP server.
feat(devcontainer): add pre-configured development container.

Removed

chore(legacy): remove legacy scenarios/resources folders.

⬆️ Back to Top## [0.0.1] - 2024-06-01

Added

feat(init): add initial repository structure.
feat(bicep): add basic Bicep templates.
feat(deploy): add PowerShell deployment scripts.
docs(copilot): add initial Copilot instructions.

⬆️ Back to Top## Version Numbering

This project uses Semantic Versioning:

0.x.y: pre-production development (current).
1.0.0: first stable production release (upcoming).
MAJOR: breaking changes to workflow or agent interfaces.
MINOR: new agents, demos, or significant feature additions.
PATCH: bug fixes, documentation improvements, and minor enhancements.

⬆️ Back to Top## Links

VERSION.md - Detailed version history
GitHub Releases

⬆️ Back to Top