Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add d-mariano/spicyclaude --skill design-docgit clone --depth 1 https://github.com/d-mariano/spicyclaudeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/d-mariano/spicyclaude/design-doc)<a href="https://agentmods.dev/skills/d-mariano/spicyclaude/design-doc"><img src="https://agentmods.dev/badge/skills/d-mariano/spicyclaude/design-doc/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/d-mariano/spicyclaude/design-doc"><img src="https://agentmods.dev/badge/skills/d-mariano/spicyclaude/design-doc.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00052 | $0.05353 |
| Opus 5 | $0.00026 | $0.02677 |
| Sonnet 5 | $0.00010 | $0.01071 |
| Haiku 4.5 | $0.00005 | $0.00535 |
Grade A, and why
design-doc scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 283 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Design Document Workflow
Produce a technical design document interactively from a PRD, supporting research, and optional technical artifacts. The output is an alignment artifact: it gets humans (and AI) agreeing on the shape of the system before code is written.
This workflow does not invent requirements. It synthesizes design from the inputs the user supplies, raises every choice that isn't already specified, and never makes silent assumptions on the user's behalf.
Inputs
In priority order:
- PRD (required) — problem statement, use cases, constraints, committed requirements. Authoritative: its decisions are baked in and not re-litigated.
- Research (recommended) — prior art, technology comparisons, recommendations, alternatives. Informational: recommendations and alternatives feed
AskUserQuestionoptions; user owns the decision. - Technical artifacts (optional) — code-as-design files: abstractions, contracts, type definitions, data models. Treat as decided for whatever they specify; cite them like a PRD section.
Input modes
- In context — PRD/research/artifacts are already in the conversation. Use what's there; don't re-read.
- On disk — user supplied paths. Read each one in full.
If you can't tell which mode applies, ask via AskUserQuestion. If no PRD exists in either form, confirm via AskUserQuestion whether to proceed in spike mode (every requirement-shaped value becomes an Open Question with a working assumption instead of a concrete commitment — for exploration, not committed design).
Critical Rules
- Don't fabricate. Every value in the design must trace back to an input or to a user decision.
- No silent assumptions. Every choice the user has not made must be either (a) sourced from PRD/research/artifacts, (b) raised via
AskUserQuestionand decided in the relevant section, or (c) recorded as deferred with a working assumption the user explicitly accepts. - Options come from given context only. Don't draw on training-data preferences. If a decision needs an option set and the inputs don't suggest any, that's an unknown — flag it, don't invent options.
- Defer requires a working assumption. A defer with no assumption is invalid; mark such questions blocking instead.
- No silent assumptions. Every choice the user has not made must be either (a) sourced from PRD/research/artifacts, (b) raised via
- Push back when something is unclear, contradictory, or risky. If a user choice conflicts with an earlier section, contradicts the PRD, or seems insane given the inputs, surface the conflict before recording.
- Stay at alignment altitude. Describe what the system is — components, contracts, data models, relationships. Code bodies, file-by-file responsibilities, build steps, and S/M/L estimates are out of scope.
- Name things concretely. Every type, interface, function, and file must have a real name. "SomeService", "DataProcessor", "handleData" are banned.
- The design file must stand alone. Anyone reading it cold must understand the system without the conversation.
- Save artifacts incrementally, never silently overwrite. Build
/context/[nnn]-<slug>/tdd-[nnn].mdsection-by-section as choices resolve. Workflow Step 0 covers what to do when files already exist. AskUserQuestionis the only sanctioned channel for design decisions. Plain-text "which would you prefer?" prompts are not allowed for any decision that lands intdd-[nnn].mdorcontracts-[nnn].md. Conversational follow-ups (e.g., when the user says "discuss further") and free-form interjections are fine.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 283 lines · 52 tokens per session scan A c9cccb4c583a
design-doc is a skill published in the GitHub repository d-mariano/spicyclaude (5 stars, last pushed 1mo ago), licensed MIT. It adds 52 tokens to every session and 5,353 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…