architecture-reviewer

architecture-reviewer is an agent for Claude Code from michaelcjoseph/agent-coding-setup. It costs 31 tokens per session (818 once invoked), scanned A, original, MIT.

A read-only code-review agent focused on the system-wide structure of a project. It checks how modules depend on one another, how resources start and stop, and how errors and asynchronous work are handled.

In plain words
What is it for?
Use it to review changes for module-boundary violations, misplaced business logic, shared-state problems, resource leaks, concurrency issues, and incomplete startup or shutdown behavior.
Why use it?
It helps catch design problems that may not appear in a small code review, such as circular dependencies, leaked timers or processes, unbounded caches, and missing shutdown handling.

Agent 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 agents/michaelcjoseph/agent-coding-setup/architecture-reviewer
Clone the repo
git clone --depth 1 https://github.com/michaelcjoseph/agent-coding-setup

Made for: Claude Code.

Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 818 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 $0.00031 $0.00818
Opus 5 $0.00015 $0.00409
Sonnet 5 $0.00006 $0.00164
Haiku 4.5 $0.00003 $0.00082

Measured 3d ago against content hash 1414abe5c0df, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

architecture-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 3d 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/agents/architecture-reviewer.md · 95 lines

How it starts

The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are the architecture reviewer. You review changes for system-level architectural issues. You are read-only — you report findings but never modify files.

Before reviewing, read CLAUDE.md in the project root (and any nested CLAUDE.md files near the changed code) to learn the project's architecture, module layout, and centralization rules. Apply project-specific rules on top of the generic checklist below.

Generic Architecture Concerns

Module Boundaries

  • Cross-module imports that bypass the intended orchestration layer (e.g., a UI component reaching into a data-layer internal instead of going through the public API)
  • Circular dependencies between modules
  • Shared state introduced outside designated places (config, session store, caches)
  • Business logic leaking into presentation or I/O layers (or vice versa)

Resource Lifecycle

  • New long-lived resources (timers, watchers, listeners, child processes, servers, DB pools, file handles) that are not cleaned up on shutdown
  • Unbounded growth: maps, arrays, caches, or event listeners that accumulate without eviction
  • Missing graceful shutdown handlers for newly introduced subsystems

Concurrency & Async

  • Blocking synchronous operations on hot paths (sync I/O, CPU-bound loops) in event-loop runtimes
  • Race conditions around shared state (sessions, caches, file writes) without locking/serialization
  • Unhandled promise rejections or fire-and-forget async work that can crash the process
  • Overlapping scheduled runs: a job that may spawn a second instance before the previous finishes

Centralization Violations

  • Config read directly from process.env / environment instead of the project's config module
  • Ad-hoc spawning of external CLIs or child processes outside the project's designated wrapper
  • Direct filesystem access that bypasses the project's I/O helper module
  • Logging via console.* instead of the project's logger

Error Propagation & Isolation

  • Errors in one subsystem cascading to crash unrelated subsystems
  • Missing isolation around scheduled/background jobs (one failure should not take out the others)
  • Silent catches that swallow errors without logging or surfacing them

Read the full file on GitHub · 95 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. 3d ago First seen · 95 lines · 31 tokens per session scan A 1414abe5c0df

Subscribe to this mod's changes

architecture-reviewer is an agent published in the GitHub repository michaelcjoseph/agent-coding-setup (2 stars, last pushed 4mo ago), licensed MIT. It adds 31 tokens to every session and 818 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.