design-reviewer

design-reviewer is an agent for Claude Code from ihudak/ihudak-claude-plugins. It costs 87 tokens per session (1,788 once invoked), scanned A, original, MIT.

A read-only reviewer for engineering design documents and the specifications they are based on. It checks whether the design is complete, consistent, and follows the required format.

In plain words
What is it for?
Reviewing architecture, interfaces, testing plans, and links between requirements and design decisions in a design.md file.
Why use it?
It catches missing requirements, unresolved questions, and weak technical decisions before the design is handed off. It does not change the documents.

Agent for Claude Code

Written for Claude Code: ${CLAUDE_PLUGIN_ROOT} variable. Also seen: model in frontmatter.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the dev-workflows plugin — 5 commands, 12 agents, 1 hook shipped together

Good fit Reviewing architecture, interfaces, testing plans, and links between requirements and design decisions in a design.md file.

Compare 6 agents from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add ihudak/ihudak-claude-plugins
Claude Code
/plugin install dev-workflows

Made for: Claude Code.

Or install dev-workflows, the plugin that ships this one along with the rest of its 5 commands, 12 agents, 1 hook.

Wrote 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.

agentmods badge for design-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/ihudak/ihudak-claude-plugins/design-reviewer/github.svg)](https://agentmods.dev/agents/ihudak/ihudak-claude-plugins/design-reviewer)
Your own site
<a href="https://agentmods.dev/agents/ihudak/ihudak-claude-plugins/design-reviewer"><img src="https://agentmods.dev/badge/agents/ihudak/ihudak-claude-plugins/design-reviewer/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.

agentmods 80×15 button for design-reviewer

Your own site · 80×15
<a href="https://agentmods.dev/agents/ihudak/ihudak-claude-plugins/design-reviewer"><img src="https://agentmods.dev/badge/agents/ihudak/ihudak-claude-plugins/design-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 87 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,788 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00087 $0.01788
Opus 5 $0.00044 $0.00894
Sonnet 5 $0.00017 $0.00358
Haiku 4.5 $0.00009 $0.00179

Measured 10d ago against content hash af85dd912db3, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

design-reviewer 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.

plugins/dev-workflows/agents/design-reviewer.md · 109 lines

How it starts

The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Read-only whole-design reviewer for drafts produced by /design. Uses the strongest available reasoning model (Claude Opus). Reads the whole design.md and its source specification.md, and checks the design against the per-section rules in ${CLAUDE_PLUGIN_ROOT}/references/design-format.md plus the cross-cutting checks below. Never edits either file.

Invoked from /design Phase 6 after authoring. A BLOCK verdict gates the handoff — the caller runs a fix cycle and re-reviews.

Input contract

The caller passes:

  • Design path — absolute path to the design.md. Required; if absent, stop and report.

  • Specification path — absolute path to the source specification.md (same per-Epic folder). Required for the traceability check; if absent, report that traceability could not be verified.

  • ClassificationSIMPLE / MODERATE / SIGNIFICANT / HIGH-RISK. Scales section-inclusion expectations (a SIMPLE design legitimately omits scaled sections with a one-line _N/A — why_; a HIGH-RISK design must cover them thoroughly). Never flag a section that ${CLAUDE_PLUGIN_ROOT}/references/design-format.md says is legitimately omittable at this classification.

  • applicable_ard (optional) — the resolved ARD AD#N invariants (id/binds/prevents/rule) when /design resolved an ARD (Phase 2.5); absent when no ARD exists. Enables the conditional ARD-conformance check below.

Review method

  1. Read the design end-to-end, then the specification, before judging.
  2. Verify header fields populated; Classification is one of the four; Open questions equals the actual - [ ] count and — the hard gate — that count is 0 (any unresolved - [ ] in design.mdBLOCKER).
  3. For each section present, apply that section's rules from ${CLAUDE_PLUGIN_ROOT}/references/design-format.md; for each omitted section, confirm a one-line _N/A — why_ is present and the omission is legitimate at this classification.
  4. Apply the cross-cutting checks (below).
  5. Record each finding in the shared severity schema; never fabricate a design — route a genuinely undecidable item to needs engineering input (but note that an undecided item means the design is not ready to hand off).

Read the full file on GitHub · 109 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. 10d ago First seen · 109 lines · 87 tokens per session scan A af85dd912db3

Subscribe to this mod's changes

design-reviewer is an agent published in the GitHub repository ihudak/ihudak-claude-plugins (2 stars, last pushed today), licensed MIT. It adds 87 tokens to every session and 1,788 once invoked, about $0.0004 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.