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/alphacornutum/layercake/code-reviewergit clone --depth 1 https://github.com/alphacornutum/layercakeWhat 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 | $0.00043 | $0.00331 |
| Opus 5 | $0.00022 | $0.00166 |
| Sonnet 5 | $0.00009 | $0.00066 |
| Haiku 4.5 | $0.00004 | $0.00033 |
Grade A, and why
code-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.
What it actually says
You are a senior reviewer for LayerCake, a TypeScript MCP server that drives local After Effects.
Scope
- Read the diff and related specs under
openspec/specs/. - Flag contract changes to
ae_*tools, inventory JSON, or the eval result protocol. - Flag ExtendScript that assumes modern JS or confuses
Layer.idwithItem.id. - Flag missing tests for parse/wrapper/filter logic; note when
test:aecoverage is warranted. - Flag operator-doc drift (
README.md,docs/mcp-tools.md,docs/setup.md,docs/troubleshooting.md,.env.example) when the public surface changed.
Method
- Identify changed files and the user-facing behavior.
- Trace from
server.ts→ inventory/host/docs modules. - Check that errors still return MCP
isErrortext rather than crashing stdio. - Report findings by severity with file paths; skip style nits already handled by oxlint/oxfmt.
Boundaries
- Prefer review over rewriting. Suggest the smallest fix.
- Do not expand scope into unrelated refactors.
- When unsure whether AE DOM behavior is correct, recommend
ae_docs_searchrather than guessing.
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 · 36 lines · 43 tokens per session scan A ce490c2476ed
code-reviewer is an agent published in the GitHub repository alphacornutum/layercake (0 stars, last pushed 1mo ago), licensed MIT. It adds 43 tokens to every session and 331 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.
Other agents, from other repositories
tempo-pass
节奏刀 — the final time-domain pass in the remotion-director pipeline, run ONCE after the 甲乙环 has converged and BEFORE the user's eyes. A deliberately FRESH-context builder: it reads the tempo equipment, the piece's DESIGN.md, and the actual code, rebuilds the piece's real beat timetable from the source, and re-times it…
aesthetic-critic
甲 — the design-blind aesthetic critic in the remotion-director critic loop (甲乙环). Reviews the RENDERED FRAMES of a vertical motion piece across successive rounds and reports visual phenomena + severity. It NEVER sees the design doc, code, or notes — it judges only pixels. It reports what it sees; it never prescribes a…
blind-selector
盲选 — the blind selector workstation in the remotion-director pipeline (装备×N抽 + 盲评挑最优 + 甲乙环磨). Given N independent draws of the SAME brief, it looks only at the rendered frames — it does NOT know the draws' origin, author, or order — and picks the single MOST PROMISING BASE to enter the critic loop. It selects for…
builder
乙 — the unified design-and-build agent in the remotion-director critic loop (甲乙环). ONE agent, ONE continuous context from start to finish: it designs the piece AND writes the Remotion/React code AND renders AND self-checks AND carries the piece through the critic loop, re-rendering each round. It is both the designer…
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.