agent-dispatch SKILL.md skillA
agent-dispatch from ninjasln-labs/agent-skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add ninjasln-labs/agent-skills/agent-dispatch@v1.5.6Pins v1.5.6 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- ninjasln-labs/agent-skills
- Path
agent-dispatch/SKILL.md- Versions recorded
- 1
- Latest version
- v1.5.6, committed Oct 5, 2026
- Size
- 13,449 bytes
- Repository stars
- 3
- License
- none declared
User-invoked only: run it when the user explicitly asks to hand work to an external agent CLI, never on your own initiative. On enable it probes the machine first — which agent CLIs are on PATH, and which LLM each one runs, read from that CLI's own --help / --list-models, never from memory or this file — then asks the user, one question at a time, to pick the agents, their model modes, and whether to append. It drives the loop after that: a task brief carrying only checked-in artifact paths…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Agent Dispatch（异构 agent 调遣） 外部 agent 是**被调遣的，不是裁定者**：合并权与闸责任始终在主会话。 本仓若另有约束本机制的决策记录（ADR 一类）⇒ **以该仓的决策记录为准**；与本件正文冲突时报用户裁决，不自行取舍。 ## 第 0 步：探测本机可支配的 agent（启用即跑） **每次启用都重探，不缓存上一轮结论**——工具升级会让旗标和模型都变，而规则 12 要求判据输入是当场落盘的命令输出，不是「谁记得」。 1. 探测起点（是搜索名单，不是事实清单）： `mcode` `command-code` `pi` `qodercn` `claude` `codex` `gemini` `cursor-agent` `aider`。 2. 逐个 `command -v <cli>`；对存在者跑 `<cli> --help`（或 `-h`）。模型面按这个梯子往下试，**试到底再判「未知」**： `--list-models`／`--models` → 模型或 provider 子命令（`models`／`provider list`，能出 JSON 的加 `--json`）→ 该 CLI 自己在用户目录下的**缓存目录文件**（如 `~/.<tool>/cache/*models*.json`，看它的 `source` 与 `updatedAt`）。 只试一条就下「未知」结论＝探测偷懒（2026-10-05 实测：`mcode` 两条旗标都不支持，但它的缓存 catalog 里模型齐全）。 3. **模型归属只认该 CLI 自己打印出来的**。打印不出、或打印的是别名而不是模型名 ⇒ 表里写「模型未知，须问用户」， 不得凭记忆、不得拿本文件或其他仓的记录当证据（那正是「二手摘要当依据」的形状）。 4. 非交互调用形同样现场复核：`-p` / `--print` / `exec` 一类，外加限轮、免落盘、结构化输出这些关键开关。 5. 判存活与判死要分清楚：无头模式常以 `node` 为进程名，按进程名 `pgrep -c` 匹配不到，须 `pgrep -f '<cli> -p'`。 6. 探测表**默认不打印**——它是过程产物，不是结论。命令必须真跑、输出真看过， 但只把结论用于 Q1 候选与注入文本；用户要核、或探测出现自相矛盾时才把表摊开。 摆出来等点将时： | agent | LLM 模型 | 模型来源 | 非交互调用形 | 已知限制 | |---|---|---|---|---| 「模型来源」只允许四种值：**现场打印**（`--help`／`--list-models`／子命令 JSON／该 CLI 的缓存目录文件，四者都算现场）／ …
Read the whole file at its exact version.
How to install
mdr add ninjasln-labs/agent-skills/agent-dispatch@sha256:5ef24b0fff7c0d7bThe line at the top of this page pins the label v1.5.6, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_unbqwcxwdl5tbpes)
2 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 1 version of this file, committed on Oct 5, 2026.
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (13449 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
ninjasln-labs/agent-skills · 3 stars · license none declared · pushed 2026-10-05 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_unbqwcxwdl5tbpes GET https://markdownregistry.com/api/v1/resolve?ref=ninjasln-labs/agent-skills/agent-dispatch GET https://markdownregistry.com/api/v1/blob/5ef24b0fff7c0d7b77b6bfc3c93df7315ae314039e3dd51ba1043b4ea80d0911
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.
More from ninjasln-labs/agent-skills
Every file in ninjasln-labs/agent-skills