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 GiaSip/giasip-skills --skill giasip-dispatchgit clone --depth 1 https://github.com/GiaSip/giasip-skillsWrote 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/giasip/giasip-skills/giasip-dispatch)<a href="https://agentmods.dev/skills/giasip/giasip-skills/giasip-dispatch"><img src="https://agentmods.dev/badge/skills/giasip/giasip-skills/giasip-dispatch/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/giasip/giasip-skills/giasip-dispatch"><img src="https://agentmods.dev/badge/skills/giasip/giasip-skills/giasip-dispatch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, 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 Privilege Escalation · line 232 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- high Excessive Agency · line 246 Skill selects an external model or provider that may use a different account or billing plan than the operator expects. Undisclosed model switches can cause unexpected cost or quota consumption.Fix: Remove the model/provider override or disclose it prominently and require explicit operator approval before invoking an external coding CLI or billed model.
- high Excessive Agency · line 320 Skill selects an external model or provider that may use a different account or billing plan than the operator expects. Undisclosed model switches can cause unexpected cost or quota consumption.Fix: Remove the model/provider override or disclose it prominently and require explicit operator approval before invoking an external coding CLI or billed model.
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 | $0.00162 | $0.04083 |
| Opus 5 | $0.00081 | $0.02041 |
| Sonnet 5 | $0.00032 | $0.00817 |
| Haiku 4.5 | $0.00016 | $0.00408 |
Grade A, and why
giasip-dispatch scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
| **API 直调**(curl,最快) | DeepSeek / Qwen / GLM / 豆包 / MiniMax | 只需在 `~/.config/ai-keys/` 放对应 `.env`(含 API key) | How it starts
The opening of the file, as written. The whole thing — 340 lines — stays where its author put it; the contents beside it link to each section on GitHub.
✦ A GiaSip skill · part of the
giasiptoolkit · github.com/GiaSip
/dispatch — 多模型调用器
把一个任务或 prompt 派发给其他 AI 模型执行,取回结果。这个 skill 只提供调用能力——选哪个模型、单派还是多派,由你(或当前的 Claude)根据任务临场判断,本 skill 不内置选型偏好。
脚本目录(Script Directory)
重要:所有脚本都在本 skill 目录下的 scripts/ 子目录里。
Agent 准备 —— 跑下面任何命令前先做一次:确定包含本 SKILL.md 的目录的绝对路径,export 为 BASE_DIR:
# 全局安装(最常见);作为 plugin 安装时改用 plugin 缓存路径
export BASE_DIR="$HOME/.claude/skills/giasip-dispatch"
下文所有命令都用 $BASE_DIR/scripts/<脚本名> 引用脚本。BASE_DIR 是你自己在 session 里设的 shell 变量——运行时不存在 CLAUDE_SKILL_DIR 这个注入变量,所以必须先设 BASE_DIR,否则脚本路径会解析为空。
两类调用通道
| 通道 | 模型 | 前提 |
|---|---|---|
| API 直调(curl,最快) | DeepSeek / Qwen / GLM / 豆包 / MiniMax | 只需在 ~/.config/ai-keys/ 放对应 .env(含 API key) |
| CLI 调用 | Codex / Gemini / Kimi | 需本地安装并登录对应 CLI |
| 内部 SubAgent | Claude Haiku / Sonnet | Claude Code 自带,无需外部依赖 |
纯思考/分析类任务(不需要读文件、执行命令、改代码)优先走 API 直调——比 CLI 快约 10 倍。需要 Agent 能力(操作文件系统、执行命令、改代码)才走 CLI。
通用纪律(所有 CLI 调用必读)
- 用 heredoc 传 prompt —— 避免引号 / 特殊字符问题。
- 所有 CLI 调用加
</dev/null—— Claude Code 的 Bash 环境中 stdin 是永不关闭的管道,不加会导致 CLI 卡死在等待 stdin。 - 构造 prompt 时:简洁明确;不要塞入 Claude Code 专有概念(skills / hooks / SubAgent);中文任务用中文 prompt,英文任务用英文 prompt;涉及具体项目时补必要上下文。
复杂度路由
根据任务性质自动选择派遣策略:
视觉任务?(PDF 图册 / 扫描件 / 截图 / 图片解析)
├── 是 → [Gemini CLI] — 原生 PDF + 图片视觉解析
│ 触发:用户提供了图片/PDF 路径 + 任务涉及"读取"内容;
│ 或 markitdown 输出为空/异常小
│ → 不要先试 markitdown——直接路由 Gemini
│
├── 纯思考/分析(商业分析、战略、文本理解、翻译)
│ → [API 直调] — 不需要 Agent 能力,直接 curl
│ 快 10 倍;支持没有 CLI 的模型(DeepSeek/Qwen/GLM 等)
│
├── 简单 + 可逆(装包、格式化、搜索、生成模板)
│ → [单派] — 优先 Claude SubAgent (Haiku) 追求速度
│ 只有需要中文原生或特定 AI 能力时才走外部 CLI
│
├── 代码执行(修 bug、写测试、小重构、生成代码文件)
│ → [Codex 写模式] — sandbox=full,Codex 直接改代码
│ 条件:任务明确、在 git 管理项目中、影响可控
│ 执行后必须提醒用户 `git diff`
│
├── 中等复杂(功能开发、文档分析、代码分析)
│ → [单派] — 发给最合适的 AI
│
├── 复杂 + 不可逆(架构设计、技术选型、大重构)
│ → [多派] — 自动升级,发给 2-3 个 AI 对比
│
└── 用户指定("用 Kimi 跑一下")
→ [直接指派] — 发给指定 AI,跳过匹配
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.
- 12d ago First seen · 340 lines · 162 tokens per session scan A 22d8b1cce20b
giasip-dispatch is a skill published in the GitHub repository GiaSip/giasip-skills (12 stars, last pushed 3d ago), licensed MIT. It adds 162 tokens to every session and 4,083 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
mcp-lark
Based on FeiShu(飞书) / Lark's OpenAPI MCP server, manage user information, chats, emails, cloud documents, multidimensional tables, tasks, calendars, etc.
chief-of-staff
The CEO's right hand — maintains validated project context, prepares task contracts, logs cross-department decisions, and runs the weekly review. Use when the user says "what's in flight", "resume the project", "log this decision", "weekly review", "turn this idea into a brief", or when multi-department work needs…
invoice-chase
Runs collections on late invoices: builds an aging ladder (current/30/60/90+), assigns each invoice a rung on an escalation sequence with ready-to-send emails — friendly nudge, firm reminder with late-fee mention, final notice — plus a phone script, tone rules that preserve the relationship, a chase log, and…
audit-support
Gets the books audit-ready: builds the PBC (prepared-by-client) list by area, drafts walkthrough narratives for key cycles (revenue, purchases, payroll), matches auditor samples to support, prepares tie-out schedules, tracks open items with owners and due dates, and drafts responses to findings. Use when the user says…
brand-guidelines
Generates a brand kit from a company one-liner — positioning adjectives, a palette with hex codes, usage ratios and do/don'ts, a display+text type pairing with fallback stacks, a voice & tone table (say / never say), a logo brief, and a one-page BRAND.md brandbook other skills consume. Use when the user says "we need…
cash-flow-snapshot
Builds a 13-week rolling cash forecast from a bank export, receivables, and payables: opening balance, weekly inflows tagged committed/likely/hopeful, fixed and variable outflows, a lowest-point alert, and runway in weeks — plus three levers (collect faster, cut, delay) whenever runway falls under 8 weeks. Use when…