design

A command that turns a use case into a reviewed design document before any code is written.

In plain words
What is it for?
Examining a use case, resolving gaps in its requirements, creating diagrams and design decisions, and saving the approved design for /implement.
Why use it?
It separates deciding how a feature should work from implementing it, while checking that the underlying requirements are clear and sound.

Command for Claude Code

Install

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.

agentmods
npx agentmods add commands/washingtonwei/project-pulse/design
Clone the repo
git clone --depth 1 https://github.com/Washingtonwei/project-pulse

Made for: Claude Code.

Per session 33 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,556 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00033 $0.03556
Opus 5 $0.00016 $0.01778
Sonnet 5 $0.00007 $0.00711
Haiku 4.5 $0.00003 $0.00356

Measured 2d ago against content hash 49a2c8712f54, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

design 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 2d 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.

.claude/commands/design.md · 78 lines

How it starts

The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/design — Turn a use case into an approved design-of-record

Take the use case named in $ARGUMENTS (e.g. UC-DOC-create-use-case) and produce its design-of-record: the Level-2 design doc for its UC area under docs/design/ (see ../../docs/design/README.md). This is the first of the two build stages — design is a deliberate, separately-reviewed stage, just like requirements. It stops before any code. /implement <UC-ID> is the second stage: it implements from the design this command leaves behind.

This stage runs in two beats (the RFC problem-then-solution rhythm): first firm the problem — challenge the use case, loop any spec fixes back until it is sound (🔬 Problem-validated in docs/traceability.md) — then design the solution against it for approval (📐 Designed). Trivial or already-sound use cases pass through both in one motion; the 🔬 state persists as a row's Build value only when a run pauses after the problem is firmed but before a solution design is approved (see Phase 4).

The use case is the contract; the linked FRs are the acceptance criteria; the glossary fixes the vocabulary. The design says only how the code will realize them — it cites the UC/FRs, never restates them.

The spec is authoritative but not infallible — challenge it. As you work the use case into a design, you will hit steps that are ambiguous or underspecified, assumptions that don't survive contact with the existing code, requirements that contradict each other or reality, and — because the docs are written by humans and age — unnecessary or incorrect constraints, out-of-date information, wrong assumptions, or an approach the spec suggests that simply isn't best practice. When that happens, ask the developer a clarifying question or push back — do not silently comply, do not silently invent the missing detail, and do not inherit a constraint just because it's written down. Name the gap or the better way, say what you'd do and why, and when it's a genuine spec defect propose the doc edit (per docs/CLAUDE.md, then /spec-build). A step with no backing FR, or a cited FR that doesn't exist, is a question to raise — not something to paper over. When in doubt, or when you see a better way, shout.

If no UC-ID is given, ask which use case to design (or list candidate use cases whose docs/traceability.md Build state is ❌ Not started — those with no design yet).

Work in phases. Stop for approval after Phase 2 before writing the design doc.

Read the full file on GitHub · 78 lines

Changes

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.

  1. 2d ago First seen · 78 lines · 33 tokens per session scan A 49a2c8712f54

Subscribe to this mod's changes

design is a command published in the GitHub repository Washingtonwei/project-pulse (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 33 tokens to every session and 3,556 once invoked, about $0.0002 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.