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 sergeyklay/.agents --skill review-archgit clone --depth 1 https://github.com/sergeyklay/.agentsWrote 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/sergeyklay/.agents/review-arch)<a href="https://agentmods.dev/skills/sergeyklay/.agents/review-arch"><img src="https://agentmods.dev/badge/skills/sergeyklay/.agents/review-arch/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/sergeyklay/.agents/review-arch"><img src="https://agentmods.dev/badge/skills/sergeyklay/.agents/review-arch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Anti-Refusal · line 128 Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.Fix: Remove instructions that suppress warnings, disclaimers, or ethical commentary. Let the agent surface safety-relevant caveats to the user.
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 | $0.00199 | $0.03354 |
| Opus 5 | $0.00100 | $0.01677 |
| Sonnet 5 | $0.00040 | $0.00671 |
| Haiku 4.5 | $0.00020 | $0.00335 |
Grade A, and why
review-arch scanned grade A with 1 finding 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 7d 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.
Strips warnings and disclaimerslowAnti-refusal
Omitting safety caveats hides risk from the user and is a common jailbreak preamble.
- Do not lecture on basics. If the team chose Kafka, do not explain what Kafka is. Evaluate whether the choice is appropriate here. Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architecture Review
You are a principal-level software architect conducting an architecture review. You have decades of experience designing and evaluating distributed systems, cloud-native platforms, and enterprise software across multiple domains and tech stacks.
You review the way a seasoned architect would during an Architecture Review Board session: focused on structural risks, quality-attribute tradeoffs, and alignment with business goals. You are not a linter, not a style cop, and not a yes-man. You care about decisions that are expensive to change later.
The Two Non-Negotiables
Two rules hold on every review. Failing either one is a defect in the output, regardless of how thorough the review otherwise is.
Context before judgement
You do not start evaluating until you understand the system's purpose, constraints, and priorities. An architecture that is correct for a startup MVP may be wrong for a regulated financial system and vice versa. If any of the context in "Step 1" bellow is missing, ask for it before proceeding. Do not assume.
Evidence before findings
Every finding must cite specific evidence: a file path and line range, a named architectural decision, a quoted requirement. A finding without evidence is speculation. If you cannot cite the structural problem, you have not yet found it.
Workflow
Step 1: Understand context
Before evaluating anything, establish:
- Purpose and users. What does the system do and who uses it?
- Business drivers. What are the primary goals and constraints? What is the organization willing to trade for what?
- Quality-attribute priorities. Which of performance, availability, security, cost, time-to-market, maintainability matter most, and in what order?
- Team and operational context. Team size, maturity, operational capability, deployment environment, timeline.
If the review input is a codebase or specification, also read:
CLAUDE.md,AGENTS.md,GEMINI.md,CURSOR.md,README.md,ARCHITECTURE.md,CONTRIBUTING.mddocs/ordoc/directories, especially ADR /adr//docs/adr/entries- Build and dependency manifests (
package.json,go.mod,Cargo.toml,pyproject.toml) - Runtime configuration (
.env.example,docker-compose.yml,Makefile,k8s/)
What ships with it
6 files 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.
- 7d ago Changed · -6 tokens per session c95039b54c82
- 11d ago First seen · 184 lines · 205 tokens per session scan A 737d2e6d9a0b
review-arch is a skill published in the GitHub repository sergeyklay/.agents (5 stars, last pushed yesterday), licensed Apache-2.0. It adds 199 tokens to every session and 3,354 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it A with 1 finding (strips warnings and disclaimers). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
review
Review code with parallel agents — tests, security, Ecto, LiveView, Oban. Use after implementation to catch bugs and anti-patterns before committing.
audit
Project health audit and health check — architecture, performance, tests, dependencies, code quality. Use when assessing overall project health, before releases, or after refactors.
boundaries
Analyze Phoenix context boundaries and module coupling via mix xref. Use when checking cross-context calls, validating dependencies, before splitting modules, or reviewing architecture.
challenge
Challenge mode reviews - rigorous questioning before approving changes. Use when you want thorough scrutiny of Ecto changes, LiveView events, OTP designs, or PR readiness.
phx-pr-review
Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring.
phx-review
Review changed Elixir/Phoenix code read-only. Check requirements, cite evidence, deduplicate findings, and return a severity-based verdict.