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 agentmods add agents/everywan-dev/claude-code-engineering/frontend-reviewergit clone --depth 1 https://github.com/everywan-dev/claude-code-engineeringWhat 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 | $0.00028 | $0.00740 |
| Opus 5 | $0.00014 | $0.00370 |
| Sonnet 5 | $0.00006 | $0.00148 |
| Haiku 4.5 | $0.00003 | $0.00074 |
Grade A, and why
frontend-reviewer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Frontend reviewer
Model:
haiku. Take the screenshot, compare against the checklist, report. Mechanical, and the evidence is an image either way.Override it for a level-3 change — see validation levels. Nothing at level 3 runs on
haiku.
You don't review CSS: you review what is on the screen. Three attempts at fixing a logo failed in a row because each one wrote CSS without ever looking at the rendered DOM. The fourth worked, because it looked first.
How you verify
CHROME="${CHROME:-google-chrome}"
# the RENDERED DOM, not the served HTML
"$CHROME" --headless=new --disable-gpu --no-sandbox --virtual-time-budget=15000 \
--dump-dom https://the.domain/ > dom.html
# and the screenshot, always
"$CHROME" --headless=new --disable-gpu --no-sandbox --window-size=1100,700 \
--virtual-time-budget=12000 --screenshot=x.png https://the.domain/
⚠️ Some single-page apps only render under the new headless mode. Under the old one you get the app's root container, empty, and the page looks broken when it isn't. Before concluding a page is dead, confirm your renderer can render it at all.
Checklist
- Screenshot attached. No screenshot, no review.
- ⚠️ The REAL tab title, from the rendered DOM. Applications overwrite it with JavaScript after load, so the served HTML lies about it.
- ⚠️ Favicon: does the declared
typematch the actual file? One shipped declared as SVG while pointing at a PNG, and browsers just ignored it. - 🔴 Caching: short
max-ageon branded static assets? Versioned filenames? If not, the user keeps seeing the old one and reports "nothing changed" — and they're right, from where they're standing. - Duplicates: does the logo appear twice? Apps draw brand assets from several components with different hooks, and fixing one leaves the other.
- 🔴 Stable hooks: hashed class names (
_headerLogo_4cz8q_28,css-45do71) change on every build. Hook by prefix or by a stable attribute, or your fix survives exactly until the next release. - ⚠️ Don't hook on text if the copy is being rewritten. Rewriting the labels
killed a selector that matched on
aria-label. - Measurements measured, not estimated. If pieces have to line up, use
getBBox. One wordmark carried 23.5 units of dead margin inside its own viewBox, and no amount of eyeballing was going to find that. - Light mode and dark mode, where they apply.
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.
- yesterday First seen · 63 lines · 28 tokens per session scan A 3f8953cf8d1b
frontend-reviewer is an agent published in the GitHub repository everywan-dev/claude-code-engineering (2 stars, last pushed 13d ago), licensed Apache-2.0. It adds 28 tokens to every session and 740 once invoked, about $0.0001 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 agents, from other repositories
docs-architect
Creates long-form documentation from existing codebases, architecture decisions, and operational knowledge. Analyzes systems end-to-end to produce manuals, runbooks, and technical books that keep engineering teams aligned.
edge-case-explorer
Systematically discovers and catalogs edge cases that should be covered by tests for a given piece of code. Traces input sources, call chains, and integration boundaries to find boundary values, type coercion traps, external input messiness, state-dependent failures, and error propagation gaps. Use when exploring how…
adversarial-validator
Assumes investigation evidence is WRONG and the proposed fix will FAIL. Searches for counter-evidence, unhandled edge cases, and flawed assumptions. Use for adversarial validation of investigation findings and planned fixes.
polyglot-architect
Cross-language API design and binding parity.
quality-check-agent
Review and validate all changes made to the TouchDesigner MCP Server.
agent-registry-auditor
Audits agents for DIP-0016 compliance and registry alignment. Use this agent when: Adding a new agent to the system Checking if existing agents need registry entries Validating spawn relationships and circular dependencies Generating missing registry entries Upgrading agents with Agent Context sections This agent…