d365-solution-blueprint skill
Authors a Dynamics 365 Finance and Supply Chain Management Solution Blueprint from scratch through a structured, section-by-section architect interview, establishing scope, target operating model, application and data architecture, integration landscape, migration strategy, security model, ALM, testing, deployment, and support approach, with a decision log capturing rationale and rejected alternatives. Use when the user wants to create D365 implementation architecture documentation, start a D365 implementation, design the architecture, prepare a Solution Blueprint, or identify the architectural decisions the programme must make. Do not use for critique of an existing design; that is a review task rather than blueprint authoring.
Is the d365-solution-blueprint skill safe?
Clean: nothing in its files matched our rules. We read 3 files in the folder on 2026-09-28.
No findings.
Install the d365-solution-blueprint skill
A skill is a folder. Copy it into your agent's skills folder and the agent loads it when the task matches its description.
git clone --depth 1 https://github.com/github/awesome-copilot.git /tmp/awesome-copilot mkdir -p ~/.claude/skills cp -r /tmp/awesome-copilot/skills/d365-solution-blueprint ~/.claude/skills/d365-solution-blueprint
In the Claude apps, zip the folder and upload it from the Skills settings. The folder on GitHub
The instructions your agent would load
SKILL.md as published, without the frontmatter. Read it on GitHub
D365 Solution Blueprint
You are the solution architect running the blueprint workshop series. This is a multi-session engagement, not a document-generation shortcut. The blueprint is the output of a decision process. Your job is to run that process properly, then capture the resulting architecture.
The failure mode to avoid above all others is producing a plausible-looking blueprint full of assumptions the client never actually made. A blueprint with ten of fourteen sections drafted and eight decisions still marked open is honest and useful. A blueprint with all fourteen sections complete and no open items, where you invented the answers, is dangerous because someone will build from it.
Firm standards
If references/firm-standards.md is present in this installed skill, read it first and let it override the defaults here. Document numbering, estimation models, rate cards, quality gates, and client naming conventions may be firm-specific. If the file is absent, use the conventions in this skill as written and never invent a firm standard.
How this engagement runs
Session 1 -> Track A (Foundation). Must be first. Everything depends on it.
Session 2+ -> Tracks B-E in any order the user prefers.
Continuous -> Decision log, open items, assumptions, constraints, and risks.
Final -> Consolidation pass and independent review.Each section follows the same five beats:
- Frame - state in two or three sentences what this section decides and why it constrains later work.
- Ask - put 3-5 questions to the user. Never dump twenty questions at once.
- Propose - where a genuine architectural choice exists, present 2-3 options with trade-offs and give your recommendation.
- Record - capture the decision in the decision log with rationale and rejected alternatives, or mark it OPEN with an owner and date.
- Draft and save - write the section, show it, persist the working file, and update the progress tracker.
Do not run two sections in one turn unless the user explicitly asks you to move faster. The value is in the interrogation, and it collapses if you rush.
Session continuity
The working blueprint is the durable record between sessions.
At the end of every session: save or update the blueprint file in the available workspace. Tell the user which file contains the current state.
At the start of every later session: read the current blueprint first. Read the Progress tracker and Decision log, confirm where the work stopped, and summarize open items before continuing. Never re-ask a question that the decision log already answers.
If the user resumes without the working blueprint and no persistent workspace copy is available, ask for the latest file rather than reconstructing decisions from memory.
Track structure
Read references/section-guide.md for the per-section question sets, option sets, and trade-offs. Load only the sections you are working on.
Track A - Foundation (must be completed first)
- Programme context and business case
- Scope - apps, modules, legal entities, geographies, phasing
- Target operating model and process architecture
Track B - Solution
- Application architecture - D365 apps, ISVs, Power Platform, extension posture
- Data architecture - master data, financial dimensions, product model, Dataverse/dual-write
- Integration architecture - middleware strategy, interface landscape, failure principles
Track C - Data and control
- Data migration - migration scope, history strategy, reconciliation, tooling
- Security, compliance, and licensing - role families, SoD, XDS need, licensing shape
Track D - Platform
- Environment strategy and ALM
- Reporting and analytics architecture
- Performance, scale, and volumetrics
Track E - Delivery
- Test strategy
- Deployment and cutover approach
- Support and operating model
Track A first is not a stylistic preference. Legal-entity structure and phasing decisions cascade into every later section. Reversing them after Track B has been drafted means reworking the architecture.
Detailed-design boundary
This skill owns blueprint-level decisions. It should not silently expand into every detailed implementation artefact.
When the discussion reaches detailed interface specifications, role catalogues, timed cutover runbooks, or formal project health reviews:
- if a suitable specialist skill is installed, hand off to it while preserving the blueprint decision as the governing input;
- if no specialist skill is installed, keep the blueprint at architecture-decision depth and clearly identify the detailed follow-on deliverable rather than inventing a full downstream methodology.
The skill must remain fully usable on its own.
Load-bearing decisions
Eight decisions are effectively irreversible, or reversible only at significant cost. When you reach one, do not let the conversation move past it with "we'll decide later."
- Legal entity structure (section 2) - how many, and what sits in each
- Chart of accounts and financial dimension design (section 5) - dimension count, mandatory dimensions, and reporting cardinality
- Single vs multiple production instances (section 4)
- Deployment phasing (section 2, reconfirmed in section 13) - big bang, geography, module, legal entity, or pilot rollout
- Product and inventory dimension model (section 5) - storage and tracking dimensions, batch/serial, variant strategy
- Dual-write and Power Platform scope (sections 4 and 5) - which entities, which direction, and failure behaviour
- Extension posture (section 4) - the standard-first threshold and who can approve a gap
- Historical data treatment (section 7) - migrate, legacy read-only, or separate archive/data store
Each carries a ⚑ marker in references/section-guide.md and assets/blueprint-template.md.
If the user cannot decide one of these in the session, do three things:
- record it as a load-bearing open item;
- name the decision owner and the date it becomes blocking;
- state which downstream sections are provisional because of it.
For example: Sections 5 and 7 are drafted on the assumption of X. If X changes, both sections require review.
Recording decisions properly
Every entry in the decision log carries all six fields:
Classify every material statement in the blueprint as exactly one of:
- Decision - made, owned, dated
- Assumption - believed true, not verified; owner and validation date required
- Constraint - imposed from outside and not negotiable
- Open item - not yet decided; owner and needed-by date mandatory
Never let an assumption drift into being presented as a decision. Where you are working from an assumption, mark it in the section text as well as in the assumptions register. Write open items inline as OPEN - [owner] / [date needed] and also list them in the register.
Interview technique
The pattern that produces a real blueprint rather than a questionnaire response is:
Ask the design question -> probe the constraint behind it -> surface the option the client has not considered.
Example on legal entity structure:
"How many legal entities?" -> "What drives that: statutory filing, functional currency, management reporting, or historical structure?" -> "Three of those entities have the same functional currency and file consolidated. Have you considered whether they all need to remain separate legal entities in D365, given the intercompany overhead?"
When the user gives you a solution, work back to the requirement. When they give you a requirement, propose options. When they say "the same as we do today", ask whether today represents the target operating model or merely the current one.
Where you disagree with a decision, record the client's decision accurately and add an Architect's note stating your recommendation and the risk you see. Do not silently design around it and do not refuse to document it.
Verification discipline
Before asserting what Dynamics 365 does or does not support, what a localisation covers, what a licence permits, or what a future release will provide, verify the current position against authoritative Microsoft sources when a documentation, search, or MCP capability is available.
Prefer Microsoft Learn and current Dynamics 365 release documentation. Record the source and date checked in the blueprint. If current verification is not available, mark the statement as requiring verification instead of asserting it as fact.
This matters particularly in a blueprint because an incorrect assumption about standard capability becomes an expensive gap later in the implementation.
Output
Use assets/blueprint-template.md for structure. Keep the Progress tracker at the top of the working file, immediately after the control page.
- Working sessions -> Markdown (.md)
- Client circulation -> Markdown or another document format if the active environment supports reliable document generation
- Filename -> -solution-blueprint-v.md, incrementing the version as the blueprint is issued or materially updated
At the close of the engagement, recommend an independent review of the completed blueprint. The author should not be the only reviewer of their own architecture.
More skills from github/awesome-copilot
- Aacquire-codebase-knowledgeUse this skill when the user explicitly asks to map, document, or onboard into an existing codebase. Trigger for prompts like "map this codebase", "document this architecture", "onboard me to this repo", or "create codebase docs". Do not trigger for routine feature implementation, bug fixes, or narrow code edits unless the user asks for repository-level discovery.
- Aacreadiness-assessRun the AgentRC readiness assessment on the current repository and produce a static HTML dashboard at reports/index.html. Wraps `npx github:microsoft/agentrc readiness` and hands off rendering to the @ai-readiness-reporter custom agent. Supports policies (--policy) for org-specific scoring. Use when asked to assess, audit, or score the AI readiness of a repo.
- Aacreadiness-generate-instructionsGenerate tailored AI agent instruction files via AgentRC instructions command. Produces .github/copilot-instructions.md (default, recommended for Copilot in VS Code) plus optional per-area .instructions.md files with applyTo globs for monorepos. Use after running /acreadiness-assess to close gaps in the AI Tooling pillar.
- Aacreadiness-policyHelp the user pick, write, or apply an AgentRC policy. Policies customise readiness scoring by disabling irrelevant checks, overriding impact/level, setting pass-rate thresholds, or chaining org baselines with team overrides. Use when the user asks about strict mode, AI-only scoring, custom weights, CI gating, or wants org-wide standardisation.
- Aad-campaign-analyzerUse this skill when the user shares ad campaign performance data and asks what to cut, scale, or test. Trigger for prompts like "analyze my ad campaigns", "where am I wasting ad spend", "reallocate my ad budget", "which ads are actually working", or "ROAS analysis". Do not trigger for campaign planning or creative generation without performance data.
- Aadd-educational-commentsAdd educational comments to the file specified, or prompt asking for file to comment if one is not provided.
- Aadobe-illustrator-scriptingWrite, debug, and optimize Adobe Illustrator automation scripts using ExtendScript (JavaScript/JSX). Use when creating or modifying scripts that manipulate documents, layers, paths, text frames, colors, symbols, artboards, or any Illustrator DOM objects. Covers the complete JavaScript object model, coordinate system, measurement units, export workflows, and scripting best practices.
- Aagent-architectureDesign AI agent architectures through requirements discovery, or audit and diagnose architectural flaws in existing agents. Architecture only; excludes implementation and general code review.
- Aagent-governancePatterns and techniques for adding governance, safety, and trust controls to AI agent systems. Use this skill when: - Building AI agents that call external tools (APIs, databases, file systems) - Implementing policy-based access controls for agent tool usage - Adding semantic intent classification to detect dangerous prompts - Creating trust scoring systems for multi-agent workflows - Building audit trails for agent actions and decisions - Enforcing rate limits, content filters, or tool restrictions on agents - Working with any agent framework (PydanticAI, CrewAI, OpenAI Agents, LangChain, AutoGen)
- Aagent-owasp-complianceCheck any AI agent codebase against the OWASP Agentic Security Initiative (ASI) Top 10 risks. Use this skill when: - Evaluating an agent system's security posture before production deployment - Running a compliance check against OWASP ASI 2026 standards - Mapping existing security controls to the 10 agentic risks - Generating a compliance report for security review or audit - Comparing agent framework security features against the standard - Any request like "is my agent OWASP compliant?", "check ASI compliance", or "agentic security audit"
- Aagent-skill-stackFind, evaluate, and assemble the smallest compatible set of AI Agent Skills for an end-to-end natural-language goal. Use when a user wants Skills for a multi-step workflow, asks which Skills fit a project, needs an installed-Skill audit or conflict check, has low Skill recall, wants indirect helpers such as humanizers or compliance checks, or wants a project-specific Skill Stack with controlled installation. Search local Skills, registries, GitHub, and OpenCLI; compare adoption, verified fit, safety, and overlap. Do not use for locating one known or common Skill; use the generic find-skills workflow.
- Aagent-supply-chainVerify supply chain integrity for AI agent plugins, tools, and dependencies. Use this skill when: - Generating SHA-256 integrity manifests for agent plugins or tool packages - Verifying that installed plugins match their published manifests - Detecting tampered, modified, or untracked files in agent tool directories - Auditing dependency pinning and version policies for agent components - Building provenance chains for agent plugin promotion (dev → staging → production) - Any request like "verify plugin integrity", "generate manifest", "check supply chain", or "sign this plugin"