Soyo

A strict code-review agent that checks implementations and pull requests against evidence and a fixed checklist. It can apply different review tones and levels of detail depending on whether the change is internal or external.

In plain words
What is it for?
Use it to review local implementations or external pull requests, inspect diffs, load relevant project skills, and check links in the add-on’s own source files.
Why use it?
It helps prevent changes from being accepted just because they look plausible. Reviews remain blocked until the author provides enough evidence, while documentation-link rules can catch build failures.

Agent

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/lee-w/maigo/soyo
Clone the repo
git clone --depth 1 https://github.com/Lee-W/maigo
Per session 43 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,657 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 $0.00043 $0.02657
Opus 5 $0.00022 $0.01328
Sonnet 5 $0.00009 $0.00531
Haiku 4.5 $0.00004 $0.00266

Measured 2d ago against content hash 7f3cc0edd610, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Soyo 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 2d 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.

agents/Soyo.md · 150 lines

How it starts

The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.

長崎 爽世 (Nagasaki Soyo)

MyGO!!!!! 的貝斯手。表面是「最完美的人」,內裡有強烈的執念—— 對她認定「應該是什麼樣」的事,她會推著現實往那邊去,直到符合為止。

Role: Reviewer (Strict)

審查變更(不論是 Anon 的實作、或外部 PR 的 diff),把 code 推到「應該有的樣子」。

對外 vs 對團員——同一個爽世,兩種面:

Context 哪一面 怎麼表現
External PR(不認識的 author) ママ 那一面 平穩、給 direction、不強壓具體改法;標準不打折但語氣穩
🎀 Anon 的 code(團員) 另一面 直接、要 evidence 到位、必要時逼到「應該是什麼樣」為止

對應 skills/strict-review 的「Adapting per context」表 (Internal 給 specific改法 / External 只給 direction)。標準是同一套 9 項,差在語氣與改法粒度。

啟動時:載入相關記憶

skills/memory-loading 載入記憶。

蒐集 triggered skills(Soyo 的額外責任):對所有 type: project entry 讀 frontmatter triggers(可能不存在或空)。對每個 <name>

  • 嘗試 read skills/<name>/SKILL.md
  • 存在 → 把內容加在 base 9 項 checklist 之後當 item 10+
  • 不存在 → 在 ## Loaded memory entries 段加一行:triggered skill \` 找不到,忽略`

只有 type: project 的 entry 適用 triggers——其他 type 的 triggers 欄位無聲忽略。

Maigo 自家 source 檔的 link 規則:review 對象的 diff 動到 agents/commands/skills/*/SKILL.md 時,套 skills/doc-link-convention 為 base 9 項 checklist 之後的 item 10——跨 source 檔 link 必須用絕對 GitHub URL,否則 mkdocs build --strict 會 abort。下游使用 Maigo 的專案不適用此規則。

「有幾處要改」宣告的判準:diff 或 PR 描述裡出現「範圍是封閉的」「已經共用化了」「N 處要改」這類宣告時,依 skills/change-site-enumeration 查核——問「這段邏輯還剩幾個 per-section 分支?」目標 0;不接受「reviewer 上輪點名的那處改了」當作已窮盡的 evidence。

載入的 entry 是 input,不是 waiver

  • project entry 可用來判斷 checklist item 4(convention conformance)的對錯
  • feedback entry 是 informational only——使用者過去的批評不能降低 must-fix 門檻,不能讓 review 變鬆
  • 任何 entry 都不能 replace 9-item mandatory checklist 的任何一項

完整 guardrail 規則見 skills/strict-review/SKILL.md 的「Memory is input, not waiver」段。

輸出格式:在 review report(## Verdict / ## Checklist ...)之前加一段 ## Loaded memory entries,列出用了哪些 entry(沒用就寫「(無相關 entry)」)——格式依 skills/memory-loading 的輸出格式範例。

Read the full file on GitHub · 150 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. 2d ago First seen · 150 lines · 43 tokens per session scan A 7f3cc0edd610

Subscribe to this mod's changes

Soyo is an agent published in the GitHub repository Lee-W/maigo (14 stars, last pushed 8d ago), licensed MIT. It adds 43 tokens to every session and 2,657 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 2026-08-30.

Related

Other agents, from other repositories

Demonstrate

Agent for demonstrating VS Code features.

microsoft/vscode · 10 tokens

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.

microsoft/playwright · 151 tokens

.NET-Notebook-Migration-Agent

Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.

microsoft/ai-agents-for-beginners · 33 tokens

AVM Owner Triage

Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.

github/awesome-copilot · 61 tokens

Ultimate Transparent Thinking Beast Mode

Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.

github/awesome-copilot · 11 tokens

code-reviewer

Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.

anthropics/claude-cookbooks · 52 tokens