| name | asset-spec |
| description | Generate artifact specifications from CDDs and related docs. Game: visual/audio/VFX asset specs and AI prompts. Product: API schema, CLI help, docs asset, config sample, migration, deployment, or package artifact specs. |
| argument-hint | [system:<name> | level:<name> | character:<name> | api:<name> | cli:<name> | sdk:<name> | docs:<name> | config:<name> | migration:<name> | deployment:<name> | package:<name>] [--review full|lean|solo] |
| user-invocable | true |
| allowed-tools | Read, Glob, Grep, Write, Edit, Task, AskUserQuestion |
User Guide
- When to use: Generate artifact specifications from CDDs and related docs. Game: visual/audio/VFX asset specs and AI prompts. Product: API schema, CLI help, docs asset, config sample, migration, deployment, or package artifact specs.
- Inputs: Command arguments:
/asset-spec [system:<name> | level:<name> | character:<name> | api:<name> | cli:<name> | sdk:<name> | docs:<name> | config:<name> | migration:<name> | deployment:<name> | package:<name>] [--review full|lean|solo]; project artifacts referenced below; user decisions and approvals before writes.
- Outputs: Primary artifacts, reports, or conversation guidance described below; write files only after user approval.
- Memory-bank writes: None.
- Next steps: Follow the workflow hand-off or next-step guidance below; recommendations do not auto-run and require explicit user command/approval.
Phase 0: Domain Routing
Detect the project domain before generating specs:
design/cdd/game-concept.md -> [Game] keep the existing visual/game asset specification workflow for sprites, meshes, VFX, audio, UI assets, levels, and engine constraints.design/cdd/product-concept.md -> [Product] use this same command for product artifact specs: API schema files, CLI help output, SDK examples, docs assets, UI screenshots, config samples, migration artifacts, and deployment/package deliverables.- If unclear, ask which artifact family is being specified.
Do not remove game asset examples. Product artifacts are an additional spec family.
Dual-Domain Parity Contract
Keep both branches inside this command:
| Area | Game branch | Product branch |
|---|
| Context reads | design/cdd/game-concept.md, target CDD/level/character doc, design/art/art-bible.md, standards/technical-preferences.md, existing asset manifest/specs | design/cdd/product-concept.md, module CDD, design/ux/interaction-patterns.md, optional design/brand/style-guide.md, optional design/design-system.md, standards/technical-preferences.md, architecture/ADR docs, existing artifact manifest/specs |
| Steps | Identify visual/audio/VFX/UI/3D assets, confirm list, generate art and technical specs, resolve art/tech conflicts, write specs and manifest | Identify API/CLI/SDK/docs/config/migration/deployment/package/UI artifacts, confirm list, generate contract and operational specs, resolve implementation/docs/deploy conflicts, write specs and manifest |
| Outputs | design/assets/specs/[target]-assets.md plus design/assets/asset-manifest.md entries | design/assets/specs/[target]-artifacts.md plus design/assets/asset-manifest.md entries for product artifacts |
| Next steps | /asset-spec [next-context], /asset-audit, /team-polish for asset quality | /asset-spec [next-artifact], /asset-audit, /test-evidence-review, /release-checklist |
If no argument is provided, check whether design/assets/asset-manifest.md exists:
- If it exists: read it, find the first context (system/level/character) with any asset at status "Needed" but no spec file written yet, and use
AskUserQuestion:
- Prompt: "The next unspecced context is [target]. Generate asset specs for it?"
- Options:
[A] Yes — spec [target] / [B] Pick a different target / [C] Stop here
- If no manifest: fail with:
"Usage: /asset-spec system:<name> — e.g., /asset-spec system:tower-defense
Or: /asset-spec level:iron-gate-fortress / /asset-spec character:frost-warden
Product examples: /asset-spec api:billing / /asset-spec cli:import / /asset-spec migration:accounts-v2
Run after the relevant concept, CDD, and domain standards are approved."
Phase 0: Parse Arguments
Extract:
- Target type:
system, level, character, api, cli, sdk, docs, config, migration, deployment, package, or ui
- Target name: the name after the colon (normalize to kebab-case)
- Review mode:
--review [full|lean|solo] if present
Mode behavior:
- Game full (default): spawn both
art-director and technical-artist in parallel
- Game lean: spawn
art-director only — faster, skips technical constraint pass
- Game solo: no agent spawning — main session writes specs from art bible rules alone. Use for simple asset categories or when speed matters more than depth
- Product full (default): spawn
lead-programmer, devops-engineer when deployment/package/migration is involved, and ux-designer when API/CLI/UI/docs consumer behavior is involved
- Product lean: spawn only the primary owner (
lead-programmer for contracts/code artifacts, devops-engineer for deployment/package/migration, ux-designer for docs/help/UI artifacts)
- Product solo: derive specs from CDD, architecture, interaction patterns, and technical preferences; flag missing specialist review in the output
Phase 1: Gather Context
Read all source material before asking the user anything.
Game Required Reads:
Product Required Reads:
- Product concept: Read
design/cdd/product-concept.md — extract core promise, users, product surface profile, MVP scope, anti-goals, and public artifact categories.
- Interaction patterns: Read
design/ux/interaction-patterns.md when the artifact affects API consumers, CLI users, SDK integrators, docs readers, or UI workflows. If missing, warn and recommend /ux-design interaction-patterns; do not fail internal headless artifacts solely for this.
- Product style references: Read
design/brand/style-guide.md if present. For UI-heavy artifacts, read design/design-system.md if present.
- Technical preferences: Read
standards/technical-preferences.md — extract language, framework, build system, package manager, naming conventions, deployment target, and test framework.
- Architecture and ADRs: Read
docs/architecture/architecture.md and relevant docs/architecture/adr-*.md files when the artifact affects contracts, deployment, migrations, auth, data, or packaging.
Source doc reads (by target type):
- system: Read
design/cdd/[target-name].md. Extract the Visual/Audio Requirements section. If it doesn't exist or reads [To be designed]:
"The Visual/Audio section of design/cdd/[target-name].md is empty. Either run /design-system [target-name] to complete the CDD, or describe the visual needs manually."
Use AskUserQuestion: [A] Describe needs manually / [B] Stop — complete the CDD first
- level: Read
design/levels/[target-name].md. Extract art requirements, asset list, VFX needs, and the art-director's production concept specs from Step 4.
- character: Read
design/narrative/characters/[target-name].md or search design/narrative/ for the character profile. Extract visual description, role, and any specified distinguishing features.
- api: Read the owning module CDD and existing
docs/api/**/*, openapi.*, schema/**/*, or API source contracts. Extract endpoint names, request/response bodies, error envelopes, auth, pagination, rate limits, idempotency, and examples.
- cli: Read the owning module CDD and existing CLI docs/source. Extract commands, flags, positional arguments, stdin/stdout/stderr behavior, prompts, exit codes, help text, config/env interactions, and examples.
- sdk: Read module CDD, architecture, package docs, and example snippets. Extract public functions/classes, typed errors, versioning, deprecation behavior, install/import shape, examples, and compatibility promises.
- docs: Read Product Concept, CDDs, UX specs, README/docs tree, and release notes. Extract pages, diagrams, screenshots, examples, tutorials, reference docs, and support/onboarding handoff.
- config: Read technical preferences, architecture, sample config files, and deployment docs. Extract required keys, defaults, validation rules, secrets policy, environment overrides, and examples.
- migration: Read data CDDs, architecture, ADRs, and existing migrations. Extract schema changes, forward/reverse behavior, dry-run requirements, data safety checks, rollback plan, and release sequencing.
- deployment: Read architecture, CI/CD config, infrastructure files, and release docs. Extract artifacts, environments, health checks, rollback, monitoring, secrets, smoke checks, and on-call handoff.
- package: Read package manifests, build config, release docs, and module CDDs. Extract package names, versioning, contents, install/upgrade behavior, signing, checksums, license files, and distribution targets.
- ui: Read UI module CDDs, UX specs,
design/ux/interaction-patterns.md, and design/design-system.md if present. Extract screens, components, states, responsive constraints, accessibility, localization, and screenshot/docs artifacts.
Optional reads:
- Existing manifest: Read
design/assets/asset-manifest.md if it exists — extract already-specced assets for this target to avoid duplicates.
- Related specs: Glob
design/assets/specs/*.md — scan for assets that could be shared (e.g., a common UI element specced for one system might apply here too).
Present context summary:
Asset Spec: [Target Type] — [Target Name]
- Source doc: [path] — [N] asset types identified
- Art bible: found — Asset Standards at Section 8
- Existing specs for this target: [N already specced / none]
- Shared assets found in other specs: [list or "none"]
Phase 2: Asset Identification
From the source doc, extract every asset type mentioned — explicit and implied.
For systems: look for VFX events, sprite references, UI elements, audio triggers, particle effects, icon needs, and any "visual feedback" language.
For levels: look for unique environment props, atmospheric VFX, lighting setups, ambient audio, skybox/background, and any area-specific materials.
For characters: look for sprite sheets (idle, walk, attack, death), portrait/avatar, VFX attached to abilities, UI representation (icon, health bar skin).
For product artifacts:
- API: schema files, request/response examples, error examples, auth docs, OpenAPI fragments, contract tests, generated clients.
- CLI: help output, command examples, shell completion, manpage/reference docs, install scripts, config samples, golden output snapshots.
- SDK/library: package metadata, typed examples, docs snippets, compatibility matrix, deprecation notes, generated docs, test fixtures.
- Docs/onboarding: diagrams, screenshots, tutorials, reference pages, release note snippets, support handoff docs.
- Config/migration/deployment/package: sample files, migration scripts, rollback notes, health checks, package manifests, checksums, release bundles.
Group assets into categories:
- Sprite / 2D Art — character sprites, UI icons, tile sheets
- VFX / Particles — hit effects, ambient particles, screen effects
- Environment — props, tiles, backgrounds, skyboxes
- UI — HUD elements, menu art, fonts (if custom)
- Audio — SFX, music tracks, ambient loops (note: audio specs are descriptions only — no generation prompts)
- 3D Assets — meshes, materials (if applicable per engine)
Present the full identified list to the user. Use AskUserQuestion:
- Prompt: "I identified [N] assets across [N] categories for [target]. Review before speccing:"
- Show the grouped list in conversation text first
- Options:
[A] Proceed — spec all of these / [B] Remove some assets / [C] Add assets I didn't catch / [D] Adjust categories
Do NOT proceed to Phase 3 without user confirmation of the asset list.
Phase 3: Spec Generation
Spawn specialist agents based on review mode. Issue all Task calls simultaneously — do not wait for one before starting the next.
Full mode — spawn in parallel:
art-director via Task:
- Provide: full asset list from Phase 2, art bible Visual Identity Statement, Color System, Shape Language, the source doc's visual requirements, and any reference games/art mentioned in the art bible Section 9
- Ask: "For each asset in this list, produce: (1) a 2–3 sentence visual description anchored to the art bible's shape language and color system — be specific enough that two different artists would produce consistent results; (2) a generation prompt ready for use with AI image tools (Midjourney/Stable Diffusion style — include style keywords, composition, color palette anchors, negative prompts); (3) which art bible rules directly govern this asset (cite by section). For audio assets, describe the sonic character instead of a generation prompt."
technical-artist via Task:
- Provide: full asset list, art bible Asset Standards (Section 8), technical-preferences.md performance budgets, engine name and version
- Ask: "For each asset in this list, specify: (1) exact dimensions or polycount (match the art bible Asset Standards tiers — do not invent new sizes); (2) file format and export settings; (3) naming convention (from technical-preferences.md); (4) any engine-specific constraints this asset type must respect; (5) LOD requirements if applicable. Flag any asset type where the art bible's preferred standard conflicts with the engine's constraints."
Lean mode — spawn art-director only (skip technical-artist).
Solo mode — skip both. Derive specs from art bible rules alone, noting that technical constraints were not validated.
Product full mode — spawn applicable owners in parallel:
lead-programmer via Task:
- Provide: product artifact list, module CDD requirements, architecture/ADR constraints, interaction patterns, technical preferences, and any existing contracts.
- Ask: "For each product artifact, produce an implementation-ready contract specification: file path, public shape, schema/help/example requirements, validation behavior, compatibility/versioning notes, tests required, and traceability to CDD/ADR requirements. Flag contradictions between docs, code, and architecture."
devops-engineer via Task when target type is deployment, package, migration, config, or release bundle:
- Provide: artifact list, deployment target, CI/CD/build system, architecture constraints, rollback requirements, and release checklist expectations.
- Ask: "For each operational artifact, specify exact file/output shape, environment assumptions, secrets policy, health/smoke checks, rollback or dry-run behavior, packaging/signing/checksum requirements, and evidence required before release."
ux-designer via Task when target type is api, cli, sdk, docs, or ui:
- Provide: interaction patterns, product concept, target user/integrator, artifact list, and docs/help requirements.
- Ask: "For each consumer-facing artifact, specify examples, help/error copy, onboarding path, empty/error/recovery states, accessibility/docs handoff, and how a user or integrator validates success."
Product lean mode — spawn only the most relevant owner for the selected target type.
Product solo mode — skip agents and derive the artifact specs from CDDs, architecture, interaction patterns, and technical preferences. Mark missing specialist review explicitly.
Collect both responses before Phase 4. If any conflict exists between art-director and technical-artist (e.g., art-director specifies 4K textures but technical-artist flags the engine budget requires 512px), surface it explicitly — do NOT silently resolve.
Phase 4: Compile and Review
Combine the agent outputs into a draft spec per asset. Present all specs in conversation text using this format:
## ASSET-[NNN] — [Asset Name]
| Field | Value |
|-------|-------|
| Category | [Sprite / VFX / Environment / UI / Audio / 3D] |
| Dimensions | [e.g. 256×256px, 4-frame sprite sheet] |
| Format | [PNG / SVG / WAV / etc.] |
| Naming | [e.g. vfx_frost_hit_01.png] |
| Polycount | [if 3D — e.g. <800 tris] |
| Texture Res | [e.g. 512px — matches Art Bible §8 Tier 2] |
**Visual Description:**
[2–3 sentences. Specific enough for two artists to produce consistent results.]
**Art Bible Anchors:**
- §3 Shape Language: [relevant rule applied]
- §4 Color System: [color role — e.g. "uses Threat Blue per semantic color rules"]
**Generation Prompt:**
[Ready-to-use prompt. Include: style keywords, composition notes, color palette anchors, lighting direction, negative prompts.]
**Status:** Needed
After presenting all specs, use AskUserQuestion:
- Prompt: "Asset specs for [target] — [N] assets. Review complete?"
- Options:
[A] Approve all — write to file / [B] Revise a specific asset / [C] Regenerate with different direction
If [B]: ask which asset and what to change. Revise inline and re-present. Do NOT re-spawn agents for minor text revisions — only re-spawn if the visual direction itself needs to change.
If [C]: ask what direction to change. Re-spawn the relevant agent with the updated brief.
For product artifacts, use this equivalent draft format:
## ARTIFACT-[NNN] — [Artifact Name]
| Field | Value |
|-------|-------|
| Category | [API schema / CLI help / SDK example / Docs asset / Config sample / Migration / Deployment / Package / UI screenshot] |
| Output Path | [expected path] |
| Owner | [lead-programmer / devops-engineer / ux-designer / docs owner] |
| Source Requirement | [CDD/ADR/story ID] |
| Validation Evidence | [contract test / CLI golden output / migration dry-run / deployment smoke / docs review] |
| Compatibility Notes | [versioning, deprecation, rollback, platform, package manager] |
**Specification:**
[Implementation-ready artifact requirements.]
**Acceptance Checks:**
- [check 1]
- [check 2]
**Status:** Needed
Phase 5: Write Spec File
After approval, ask:
- Game: "May I write the spec to
design/assets/specs/[target-name]-assets.md?"
- Product: "May I write the spec to
design/assets/specs/[target-name]-artifacts.md?"
Write the file with:
# Asset Specs — [Target Type]: [Target Name]
> **Source**: [path to source GDD/level/character doc]
> **Art Bible**: design/art/art-bible.md
> **Generated**: [date]
> **Status**: [N] assets specced / [N] approved / [N] in production / [N] done
[all asset specs in ASSET-NNN format]
For product projects, write:
# Product Artifact Specs — [Target Type]: [Target Name]
> **Source**: [path to source CDD/ADR/UX/release doc]
> **Interaction Patterns**: design/ux/interaction-patterns.md if applicable
> **Technical Preferences**: standards/technical-preferences.md
> **Generated**: [date]
> **Status**: [N] artifacts specced / [N] approved / [N] implemented / [N] verified
[all artifact specs in ARTIFACT-NNN format]
Then update design/assets/asset-manifest.md. If it doesn't exist, create it:
# Asset Manifest
> Last updated: [date]
## Progress Summary
| Total | Needed | In Progress | Done | Approved |
|-------|--------|-------------|------|----------|
| [N] | [N] | [N] | [N] | [N] |
## Assets by Context
### [Target Type]: [Target Name]
| Asset ID | Name | Category | Status | Spec File |
|----------|------|----------|--------|-----------|
| ASSET-001 | [name] | [category] | Needed | design/assets/specs/[target]-assets.md |
If the manifest already exists, append the new context block and update the Progress Summary counts.
Ask: "May I update design/assets/asset-manifest.md?"
Phase 6: Close
Use AskUserQuestion:
- Prompt: "Asset specs complete for [target]. What's next?"
- Options:
[A] Spec another system — /asset-spec system:[next-system][B] Spec a level — /asset-spec level:[level-name][C] Spec a character — /asset-spec character:[character-name][D] Spec a product artifact — /asset-spec api:[name] or /asset-spec cli:[name][E] Run /asset-audit — validate delivered assets/artifacts against specs[F] Run /test-evidence-review — verify contract, migration, package, or docs evidence[G] Stop here
Asset ID Assignment
Asset IDs are assigned sequentially across the entire project — not per-context. Read the manifest before assigning IDs to find the current highest number:
Grep pattern="ASSET-" path="design/assets/asset-manifest.md"
Start new assets from ASSET-[highest + 1]. This ensures IDs are stable and unique across the whole project.
If no manifest exists yet, start from ASSET-001.
Shared Asset Protocol
Before speccing an asset, check if an equivalent already exists in another context's spec:
- Common UI elements (health bars, score displays) are often shared across systems
- Generic environment props may appear in multiple levels
- Character VFX (hit sparks, death effects) may reuse a base spec with color variants
If a match is found: reference the existing ASSET-ID rather than creating a duplicate. Note the shared usage in the manifest's referenced-by column.
"ASSET-012 (Generic Hit Spark) already specced for Combat system. Reusing for Tower Defense — adding tower-defense to referenced-by."
Error Recovery Protocol
If any spawned agent returns BLOCKED or cannot complete:
- Surface immediately: "[AgentName]: BLOCKED — [reason]"
- In
lean mode or if technical-artist blocks: proceed with art-director output only — note that technical constraints were not validated
- In
solo mode or if art-director blocks: derive descriptions from art bible rules — flag as "Art director not consulted — verify against art bible before production"
- Always produce a partial spec — never discard work because one agent blocked
Collaborative Protocol
Every phase follows: Identify → Confirm → Generate → Review → Approve → Write
- Never spec assets without first confirming the asset list with the user
- Always anchor game specs to the art bible — a game asset spec that contradicts the art bible is wrong
- Always anchor product specs to the Product Concept, interaction patterns, architecture/ADRs, and technical preferences — a product artifact spec that contradicts those sources is wrong
- Surface all agent disagreements — do not silently pick one
- Write the spec file only after explicit approval
- Update the manifest immediately after writing the spec
Recommended Next Steps
- Run
/asset-spec [next-context] to continue speccing remaining systems, levels, or characters
- Run
/asset-spec api:[name], /asset-spec cli:[name], /asset-spec migration:[name], or another product target to continue product artifact coverage
- Run
/asset-audit to validate delivered assets/artifacts against the written specs and identify gaps or mismatches
- Run
/test-evidence-review for product contract, migration, package, deployment, and docs evidence before release
