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 instructions/mythkiven/figma-ios-codegen/agents-mdgit clone --depth 1 https://github.com/mythkiven/figma-ios-codegenWrote 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/instructions/mythkiven/figma-ios-codegen/agents-md)<a href="https://agentmods.dev/instructions/mythkiven/figma-ios-codegen/agents-md"><img src="https://agentmods.dev/badge/instructions/mythkiven/figma-ios-codegen/agents-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.01739 | $0.01739 |
| Opus 5 | $0.00870 | $0.00870 |
| Sonnet 5 | $0.00348 | $0.00348 |
| Haiku 4.5 | $0.00174 | $0.00174 |
Grade A, and why
figma-ios-codegen AGENTS.md 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.
How it starts
The opening of the file, as written. The whole thing — 133 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Project Agent Rules
本文件是跨客户端的 Agent 入口(Cursor / Claude Code / Codex / 其他读
AGENTS.md的工具)。
Cursor 细则:.cursor/rules/figma-ios-codegen-workflow.mdc(按描述触发,alwaysApply: false;命中 Figma→代码意图时加载)。
新增 / 删除 / 重命名.cursor/rules/*.mdc时,必须同步更新本文件。
给人读的对外文档(安装 / 架构 / 设计约定 / FAQ):docs/— Agent 生成代码时仍以本文件 + skills 为准,不必整份加载docs/。
必读规则文件
| 文件 | 适用场景 |
|---|---|
.cursor/rules/figma-ios-codegen-workflow.mdc |
用户给出 Figma 链接并要求转/生成 iOS 代码时:强制按该 workflow 执行 |
详细 skill:.cursor/skills/(总入口 figma-ios-playbook)。
宿主定制:.cursor/bindings/(bindings/README.md)。
当前示例前缀 MK;换宿主只改 JSON,不改 skill。
何时启用本流程
同时满足才启用:
- 提到
figma或给出figma.com/...链接 - 表达「转/生成代码」意图
不触发:只问 Figma 用法、或未提 Figma 的「写个登录页」。
当前只支持 iOS Swift(UIKit)。
铁律(最高优先级)
- 先数据包,后写码 — 无成功的
{data_dir}/manifest.json→ 禁止阶段 2。已有data_dir可跳过重拉,仍须校验schema_version。 - 阶段 2 唯一数据源 — 只读数据包 +
.cursor/bindings/;禁止再调 Figma REST / MCP。 - 全量还原 — 应实现节点全部生成;未完成须按 mdc「未完成处置模板」开头声明。
- TODO — 仅
// TODO(阶段3):/// TODO(阶段2):。 - 不要改 preload 源码「自愈」 — 先
python3 -m src.cli --self-check-only;禁止 LLM 重写src/**。
架构(1 → 2 → 4;阶段 3 仓外)
| 阶段 | 做什么 | 谁做 |
|---|---|---|
| 1 | Figma URL → 规范化数据包(零 LLM) | 本地 CLI(或可选自建 MCP) |
| 2 | 数据包 + bindings → 基础 UI Swift | Agent + .cursor/skills |
| 4 | 节点覆盖率 + 截图视觉验收 | Agent(阶段 2 之后强制) |
| 3 | PRD + 接口 → 商用业务代码 | 不在本仓库流程内 |
阶段 1:生成数据包
路径探测
WORKSPACE_ROOT="$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
PRELOAD_DIR_ENV="${FIGMA_PRELOAD_DIR:-$FIGMA_DATA_MCP_DIR}" # 新名优先;旧名兼容
if [ -n "$PRELOAD_DIR_ENV" ] && [ -f "$PRELOAD_DIR_ENV/src/cli.py" ]; then
DATA_PRELOAD_DIR="$PRELOAD_DIR_ENV"
elif [ -f "$WORKSPACE_ROOT/.cursor/figma-ios-preload-data/src/cli.py" ]; then
DATA_PRELOAD_DIR="$WORKSPACE_ROOT/.cursor/figma-ios-preload-data"
elif [ -f "$WORKSPACE_ROOT/figma-ios-preload-data/src/cli.py" ]; then
DATA_PRELOAD_DIR="$WORKSPACE_ROOT/figma-ios-preload-data"
fi
# 有 DATA_PRELOAD_DIR → MODE=local;否则需已配置 fetch_figma_design(MODE=mcp)
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.
- 3d ago First seen · 133 lines · 1,739 tokens per session scan A a93f048fe62f
figma-ios-codegen AGENTS.md is an instructions file published in the GitHub repository mythkiven/figma-ios-codegen (9 stars, last pushed 1mo ago), licensed MIT. It adds 1,739 tokens to every session, about $0.0087 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 instructions, from other repositories
figwright CLAUDE.md
Instructions for awdr74100/figwright: Read AGENTS.md first — it's the single source of guidance for this repo: architecture, layout, tech stack, commands, conventions, release flow, and gotchas.
ios-agent-skill AGENTS.md
Instructions for Nagarjuna2997/ios-agent-skill, covering ios agent skill — claude ai expert ios/swift developer, when to load this skill, loading the right document, how these docs are structured and how you operate: delegation, loops, and verification.
platform AGENTS.md
Instructions for qorm/platform, covering agents.md — orienting an ai agent in the qorm repo, understand it, write / edit an app, drive a live app as an agent and build & test.
figbridge AGENTS.md
Instructions for rudraptpsingh/figbridge, covering figbridge — agent memory, repo shape, release flow — do not deviate, mcp registry listing and seamless auto-updates for users.
flato-design-mcp AGENTS.md
Instructions for sologovision/flato-design-mcp: This repository is the public distribution package for Flato Design MCP. It contains setup docs, guide skills, Codex plugin packaging, examples, screenshots, and registry/review metadata. It does not contain the hosted MCP server source code.
flato-design-mcp CLAUDE.md
Instructions for sologovision/flato-design-mcp: Use AGENTS.md as the root launcher, then follow .harness/index.md.