Agent skill

Design Blueprint

by techygarg in techygarg/lattice

Run a complete design workflow -- from establishing context through four progressive design levels (Capabilities, Components, Interactions, Contracts) to an approved blueprint.

MITAuto-check passedDevelopment

Install Design Blueprint

skills CLI
$ npx skills add techygarg/lattice --skill design-blueprint -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install techygarg/lattice design-blueprint --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/techygarg/lattice.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/design-blueprint .claude/skills/design-blueprint && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
design-blueprint
GitHub stars
199
Token cost
~3.1k tokens
SKILL.md length
1,595 words
Files
1
Skills in repo
33
Repo updated
First seen
Licence
MIT

At a glance

Run a complete design workflow -- from establishing context through four progressive design levels (Capabilities, Components, Interactions, Contracts) to an approved blueprint.

  • Works in 3 steps: Establish Context → Walk the Design Levels → Finalize Blueprint
  • Starting a design
  • SKILL.md covers Required Skills and Workflow
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Design Blueprint is an agent skill from techygarg/lattice. Run a complete design workflow -- from establishing context through four progressive design levels (Capabilities, Components, Interactions, Contracts) to an approved blueprint. Composes knowledge-priming, context-anchoring, learning-harvest, collaborative-judgment, design-first, architecture, and domain-driven-design into one process. Handles both new features (create context doc) and resuming existing work (load context doc). Level 5 (Implementation) is delegated to code-forge. Use when starting a design…

Its SKILL.md is about 3.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Development, covering Domain-driven design. The repository describes itself as: Install engineering discipline into any AI coding assistant. Composable skills for design, implementation, review, and team standards. Better process, not just better prompts. The licence is MIT.

When your agent uses it

  • Starting a design
  • Planning architecture
  • The user says design a feature
  • Start designing

Example prompts

  • “design a feature”
  • “blueprint”
  • “start designing”
  • “/design-blueprint”

Workflow steps

3 steps, taken from the step headings in SKILL.md.

  1. Establish Context
  2. Walk the Design Levels
  3. Finalize Blueprint

What it can do on your machine

Read from SKILL.md and the folder at commit 4d6c35f. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Design Blueprint loads about 3.1k tokens when it runs. Until then it costs about 171 tokens; SKILL.md has 1,595 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~171
When it runs · the whole SKILL.md, loaded when a task matches
~3.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from techygarg/lattice at commit 4d6c35f, republished under its MIT licence (© techygarg). 1,595 words, ~3,130 tokens.

Download SKILL.mdSave it as .claude/skills/design-blueprint/SKILL.md (or your agent's skills folder).
name
design-blueprint
description
Run a complete design workflow -- from establishing context through four progressive design levels (Capabilities, Components, Interactions, Contracts) to an approved blueprint. Composes knowledge-priming, context-anchoring, learning-harvest, collaborative-judgment, design-first, architecture, and domain-driven-design into one process. Handles both new features (create context doc) and resuming existing work (load context doc). Level 5 (Implementation) is delegated to code-forge. Use when starting a design, planning architecture, or when the user says 'design a feature', 'blueprint', 'start designing', 'plan the architecture', or 'let's design before coding'.

Design Blueprint

Required Skills

Read and apply in order before Step 1:

  1. framework:knowledge-priming -- Load the project knowledge base so every decision grounds in the real project. (always)
  2. framework:context-anchoring -- Create or load the feature's living context doc (Create / Load / Enrich behaviors). (always)
  3. framework:learning-harvest -- Load prior operational learnings at session start; harvest new ones at session end. (always)
  4. framework:collaborative-judgment -- Surface genuine judgment calls as structured options instead of silently assuming. (always)
  5. framework:design-first -- Owns the 5-level methodology. Its Entry Assessment, Complexity Calibration, Simplicity Check, and Level Completion Protocol govern Step 2. (Step 2)
  6. framework:architecture -- Validate components, layers, dependency direction, and boundary rules (design mode). (Levels 2-4)
  7. framework:domain-driven-design -- Model aggregates, entities, value objects, events, and contracts (design mode). (Levels 2-4)

Workflow

Step 1: Establish Context
  1. Run framework:learning-harvest Load behavior. Focus hint: "design session — focus: design patterns, reliability, structural health".

  2. Set up the feature's living doc with framework:context-anchoring:

    • Discover: scan .lattice/context/ for an existing anchor doc matching the feature name or frontmatter.
    • Found → Load behavior. Present the structured acknowledgment: feature name, status, decision count, open questions, constraints. Then run the resume check below.
    • Not found → Create behavior. Confirm the feature name, summary, and requirement doc link with the user before creating. Then begin Step 2 — the Entry Assessment sets the entry level.
  3. Resume check (when a doc was found) — derive the earliest incomplete step from the doc itself. STOP: Never re-walk agreed work:

    • status: approved → design is finished. Say so and stop; suggest /code-forge.
    • No sections starting ## Design: Level → begin Step 2 — the Entry Assessment (and its spec rule) sets the entry level.
    • Some levels persisted → summarize the approved levels briefly, then resume at the first missing level at or after the recorded entry level (the [Entry] Decisions Log entry; older docs without one → treat entry as Level 1).
    • Every level from entry through Level 4 persisted, but no ## Design Summary, or status ≠ approved → go directly to Step 3.
  4. Requirement constraints: read requirement_doc from the context doc frontmatter.

    • Absent → skip.
    • Local path, unreadable → STOP: "Requirement doc not found at [path]. Verify before continuing."
    • Local path, readable → read it and extract ## Technical Constraints. Treat as non-negotiable — same authority as architecture rules. Surface to the user before the first level is presented.
    • External reference (URL, ticket ID, or other non-local-path identifier) → resolve via a connected MCP tool if one can. If none is connected or the fetch returns nothing, ask the user to paste the current constraints — expected, not an error.
    • Either path resolved to the spec itself (not pasted constraints only) → Step 2's spec rule applies.
    • Conflict during design → surface via framework:collaborative-judgment. The user decides; record the change back in the requirement doc's ## Technical Constraints if local, or in the Decisions Log if external — this molecule never writes to an external system.
  5. Write the back-link: if requirement_doc resolved to a readable local file at .lattice/requirements/features/{feature-name}.md, add to its ## Links section: - Design: [{feature-name}.md](../../context/{feature-name}.md). One discrete file edit; skip if the link is already present.

Step 2: Walk the Design Levels

Run design-first's Entry Assessment first: state the proposed entry level from its Complexity Calibration table and wait for confirmation. Record the confirmed entry level as the first Decisions Log entry: [Entry] Start at Level N (name) — rationale. If key use cases or success criteria are unclear, surface them via framework:collaborative-judgment before producing the first level output.

Spec as Level 1: when the context doc has no ## Design: Level section yet and requirement_doc resolved to the spec itself, the spec is Level 1. Do not re-validate it — spec quality belongs to requirement-forge.

  • Propose entry at Level 2 or the calibrated level, whichever is later, and offer: "Level 1 comes from the linked spec. Starting at Level [N]. Want a capabilities walk-through first?" STOP — do NOT advance until the user answers.
  • User wants the walk-through → enter at Level 1, using the spec as input.
  • Otherwise, save ## Design: Level 1 -- Capabilities as one line: Taken from [{spec-file}](../requirements/features/{spec-file}). Not re-walked. — the link is relative to the context doc; an external spec gets its reference instead. Do not copy the spec.

Drive the levels sequentially from the confirmed entry level through Level 4 via framework:design-first. Complexity Calibration sets how deep each level goes; it never removes a gate or skips persistence.

Gate (every level) — follow design-first's Level Completion Protocol: present the level output with its targeted gating question, then STOP — do NOT advance until the user explicitly confirms, not on silence, not on ambiguity.

Persist (after every approval, before advancing) — use framework:context-anchoring Enrich to write into the context doc:

  1. The approved output as a clean structured summary under ## Design: Level N -- {Name}, same format as presented (numbered list L1; component table + diagram L2; sequence/flow L3; typed interfaces L4). Persist diagrams as Mermaid.
  2. One Decisions Log entry per decision: [Level N] Chose X because Y. Rejected: Z.
  3. Constraints identified during the discussion (non-negotiable boundaries that emerged).
  4. Open questions surfaced but unresolved.

STOP: Do not present the next level until these writes are done.

Judgment calls: when applying architectural atoms at any level, surface genuine design judgment calls immediately via framework:collaborative-judgment — never batch them to the end of a level.

Evidence rule (Level 2): before presenting components, quickly explore the codebase and map each proposed component to the existing modules/packages it extends, wraps, or modifies — or mark it new. Present the mapping with the components. Never invent a parallel structure that ignores what exists.

Level-specific applications:

  • Level 1 (Capabilities): numbered user-facing capabilities, max 5, no technical detail (per design-first).
  • Level 2 (Components): challenge each component before approving — does it need to exist? One known implementation, one caller, or an unconfirmed problem → inline it or defer. Then validate in design mode: framework:architecture (layer mapping, dependency direction, boundary clarity) and framework:domain-driven-design (aggregates, entities, value objects; domain vs infrastructure placement).
  • Level 3 (Interactions): framework:architecture — data flows follow the loaded patterns; boundary-crossing rules respected. framework:domain-driven-design — cross-aggregate communication uses domain events / eventual consistency.
  • Level 4 (Contracts): framework:domain-driven-design — repository interfaces, value object types, aggregate root boundaries reflecting the tactical choices from earlier levels. framework:architecture — boundary-data rules and interface ownership respected. Every Level 3 interaction maps to at least one interface.

Regression rule: if the user reopens an approved level, re-run that level's gate. On re-approval, mark every downstream persisted level section stale ("stale — pending re-approval after Level N change") and re-present them for confirmation before Step 3. STOP: Never leave contradictory approved sections in the doc.

Early exit: if the user wants to stop or shortcut the design, follow design-first's Mid-level exit. Persist whatever was approved and leave status as draft — a partial doc is a valid outcome.

Show full SKILL.md (491 more words)Show less
Step 3: Finalize Blueprint

After Level 4 is approved and persisted:

  1. Verify completeness and consistency: the context doc must contain all four level sections plus every decision made during the design. Enrich anything missing now. Then check:

    • Every Level 3 interaction maps to at least one Level 4 interface.
    • Every Level 4 interface is owned by exactly one Level 2 component (a shared type is owned by its defining component). Fix any gap through the affected level's gate — never silently.
  2. Check requirement spec drift: read requirement_doc from the context doc frontmatter.

    • Absent → note in Design Summary: "No requirement doc — drift check skipped."
    • Local path, unreadable → STOP: "Requirement doc not found at [path]. Verify before continuing." (A broken local path is an error.)
    • External reference, unresolvable (no connected MCP tool, or the fetch returns nothing) → do not STOP — expected, not broken. Ask the user to paste current constraints/scenarios if a comparison is wanted, or note in Design Summary: "Requirement doc is external and unavailable this session — drift check skipped."
    • Resolved (local file read, external fetch succeeded, or user pasted constraints) → compare L4 contracts against Scenarios/ACs and ## Technical Constraints. Present each divergence as [field/behavior] — changed from [X] to [Y]. Reason: [from Decisions Log], or "L4 consistent with requirement spec — no overrides" if none. Ask: "Record this in the requirement doc?"
    • STOP: do not write to requirement_doc until confirmed. Confirmed and local → write each finding into the requirement doc's ## Links section as - Design override: [field/behavior] — changed from [X] to [Y]. Reason: [...], or - Design alignment: L4 consistent with requirement spec — no overrides. if none. Confirmed and external → this molecule never writes to an external system; record the findings in the Design Summary instead. Declined → note in Design Summary: "Drift check results not written to requirement doc — see Decisions Log."
  3. Write the design summary: use framework:context-anchoring Enrich to add a ## Design Summary section containing components and layer assignments, key contracts and interfaces, architectural constraints, domain model decisions (if applicable), and open questions resolved during design.

  4. Set approved status: write status: approved into the context doc frontmatter. STOP: discrete file edit — not prose. Without it, code-forge will not proceed. STOP: never write status to requirement_doc — the requirement's status belongs to whoever manages it (a human, or an external system); this molecule manages only its own context doc.

  5. Log the completion decision: "Design approved at Level 4. Status set to approved — ready for implementation." Present the summary to the user as confirmation.

  6. Harvest learnings: run framework:learning-harvest Harvest behavior. Session context: "design session — architectural decomposition and contract definition". Synthesize and propose cross-cutting patterns from this session — decomposition approaches, architectural trade-offs, scope decisions that could inform future designs. The user confirms what enters the document. STOP: run this before the next bullet — do not jump straight to the /code-forge suggestion.

  7. Design complete. Do NOT proceed to Level 5 (Implementation). Suggest the user invoke /code-forge when ready to begin coding against the approved design.

© techygarg, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/design-blueprint of techygarg/lattice.

Open the folder on GitHubat commit 4d6c35f

Compare with similar skills

Design Blueprint next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Design Blueprint compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Design Blueprint this skilltechygarg/lattice199—~3.1kAutomated safety check: PassMIT
Domain Modelingfossasia/eventyay-interpretation1.6k31 repos~821Automated safety check: PassApache-2.0
Architecture Governancezai-org/ZCode7.6k—~1.2kAutomated safety check: PassApache-2.0
Evolutionary Modular Architecturetech-leads-club/agent-skills7k—~3.7kAutomated safety check: PassCC-BY-4.0
Domain Modelingbrim-borium/spotify_sdk1665 repos~806Automated safety check: PassApache-2.0
Domain Modeling and Glossarywindmill-labs/windmill18k—~622Automated safety check: PassCustom licence

Similar skills

  • Domain Modeling

    fossasia/eventyay-interpretation

    Build and sharpen a project's domain model. An agent skill from fossasia/eventyay-interpretation.

    1.6k GitHub starsUsed in 31 repos~821 tokens
    DevelopmentAuto-check passed
  • Apply the repository's architecture policy to code changes by generating a bounded context package, checking module and layer boundaries, and reporting baseline-aware violations.

    7.6k GitHub stars~1.2k tokensUpdated 10 days ago
    DevelopmentAuto-check passed
  • Evolutionary Modular Architecture

    tech-leads-club/agent-skills

    Guides design of modular-monolith platforms with DDD, flat-by-aggregate modules, anti-corruption layers, outbox events and resilience, plus an architecture document with SVG diagrams.

    7k GitHub stars~3.7k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Domain Modeling

    brim-borium/spotify_sdk

    Build and sharpen a project's domain model. An agent skill from brim-borium/spotify_sdk.

    166 GitHub starsUsed in 5 repos~806 tokens
    DevelopmentAuto-check passed
  • Domain Modeling and Glossary

    windmill-labs/windmill

    Actively challenges vague or conflicting terminology as you design, and keeps a living domain glossary file up to date in real time.

    18k GitHub stars~622 tokensUpdated today
    DevelopmentAuto-check passed
  • Ddd

    swamp-club/swamp

    Domain Driven Design guidance for TypeScript/Deno codebases.

    644 GitHub stars~1.4k tokensUpdated yesterday
    DevelopmentAuto-check passed

More from techygarg/lattice

All 33 skills in this repo
  • Architecture Compass

    techygarg/lattice

    Architectural thinking partner for an existing repository — scans the codebase, conducts a structured interview, agrees on current architectural state and recommended direction, and produces a…

    199 GitHub stars~4.4k tokensUpdated 4 days ago
    Auto-check passed
  • Lattice Init

    techygarg/lattice

    Guided setup and upgrade-check experience for Lattice projects -- scans the repository, detects existing configuration and outdated conventions, suggests refiners and available upgrades in priority…

    199 GitHub stars~3.3k tokensUpdated 4 days ago
    Auto-check passed
  • Skill Align

    techygarg/lattice

    Audit and fix all Lattice documentation, README, docs/, PROJECT.md, GitHub issue templates, and CLAUDE.md to ensure they are fully aligned with the current skill inventory.

    199 GitHub stars~2k tokensUpdated 4 days ago
    Auto-check passed
  • Skill Validate

    techygarg/lattice

    Validate any Lattice SKILL.md against all tier conventions — atoms, molecules, and refiners.

    199 GitHub stars~1.5k tokensUpdated 4 days ago
    Auto-check passed
  • Architecture Refiner

    techygarg/lattice

    Facilitate a structured conversation to define architecture principles for a repository.

    199 GitHub stars~3.7k tokensUpdated 4 days ago
    Auto-check passed
  • Clean Code Refiner

    techygarg/lattice

    Facilitate a structured conversation to define clean code principles for a repository.

    199 GitHub stars~3k tokensUpdated 4 days ago
    Auto-check passed

Categories

Questions about Design Blueprint

What does Design Blueprint do?

Run a complete design workflow -- from establishing context through four progressive design levels (Capabilities, Components, Interactions, Contracts) to an approved blueprint. Design Blueprint is an agent skill from techygarg/lattice. Run a complete design workflow -- from establishing context through four progressive design levels (Capabilities, Components, Interactions, Contracts) to an approved blueprint.

When should I use Design Blueprint?

Design Blueprint fits situations like: starting a design; planning architecture; the user says design a feature; start designing.

How do I install Design Blueprint in Claude Code?

Run `npx skills add techygarg/lattice --skill design-blueprint -a claude-code`. Or copy the skill folder (skills/design-blueprint in techygarg/lattice) into .claude/skills/design-blueprint in your project. Claude Code loads it when a task matches its description.

How do I install Design Blueprint in Codex?

Run `npx skills add techygarg/lattice --skill design-blueprint -a codex`. Or copy the skill folder (skills/design-blueprint in techygarg/lattice) into .agents/skills/design-blueprint in your project. Codex loads it when a task matches its description.

Can I use Design Blueprint in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add techygarg/lattice --skill design-blueprint -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/design-blueprint, .gemini/skills/design-blueprint, .github/skills/design-blueprint and .opencode/skills/design-blueprint in your project.

What does Design Blueprint need to run?

SKILL.md names no scripts, command-line tools or credentials: Design Blueprint is instructions for the agent only.

Does Design Blueprint access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Design Blueprint safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Design Blueprint use?

Design Blueprint is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Design Blueprint use?

About 3.1k tokens (SKILL.md is roughly 13k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Design Blueprint?

Skills that share tags, products or a category with Design Blueprint: Domain Modeling (fossasia/eventyay-interpretation, 1.6k stars), Architecture Governance (zai-org/ZCode, 7.6k stars), Evolutionary Modular Architecture (tech-leads-club/agent-skills, 7k stars) and Domain Modeling (brim-borium/spotify_sdk, 166 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Design Blueprint?

techygarg (a GitHub user) maintains it in techygarg/lattice, which has 199 GitHub stars. The repository holds 33 skills in this directory. The repository was last updated on October 6, 2026.

Source: techygarg/lattice on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.