All Claude Code modsalexcz-a11y/claude-mods
dispatch-pilot
Dispatch Pilot:一个在 Claude 之外的决策模型(默认 Perplexity 的 pplx,没有 Perplexity key 时退回 TypeSafe 的 Jev,也可以固定用 Jev)在你每次发消息时决定主 agent 这一轮的 effort,每一步都按这一档发出,一轮进行中每隔几步再判断一次;也为主 agent 派出的每个 agent 和 Workflow 脚本里的每个 agent 按 Artificial Analysis 的基准选模型和 effort(sonnet 和 opus 至少 medium)。主 agent 的模型从不改变,prompt cache 不受影响(只在 Claude Code 订阅下成立;Bedrock、Vertex 和网关上切换 effort 会让缓存失效)。它还不让主 agent 读完整的 skill 列表,改为在你发消息时推荐合适的几个,一轮中途主 agent 可以用 find_skill 工具查 skill。它还能替 agent 看图核对:agent 交上截图或录屏和几条检查项,由 TypeSafe 的 Jev(没有 TypeSafe key 时用 pplx)一两秒内回答过、不过或没把握,不必自己看;可选的截图接管把 agent 自己生成的图换成一句说明,带图路由按参考图的复杂度定 effort(这两项默认关。图没法脱敏:带图路由会把你贴的参考图发给 TypeSafe 或 Perplexity,看图核对会把 agent 交上来的图发过去;截图接管自己不发图)。agent 的工具调用失败累计到一定次数时强制升高它的 effort(haiku agent 的 effort 到顶后改用 sonnet),除非这些失败本来就在意料之中。所有判断和理由都放在看板上:prompt 上方的 band 列出本轮每个 agent 的模型和 effort,脚部一个短摘要,/dp 打开的依据面板写明每个决定的依据和决策日志。
review UI rewrite
The scan could not record a complete footprint for dispatch-pilot, so its access is unknown; it passed claude plugin validate on Claude Code 2.1.302.
- Access: unknown, the footprint is incomplete
- Static scan, not a runtime test
Install hint
/plugin install dispatch-pilot --marketplace alexcz-a11y/claude-modsFrom the repository's own marketplace. Confirm the plugin name in the repository's README before installing. Mods need Claude Code 2.1.287 or later.
Access
Unknown accessThe footprint is unavailable or incomplete.Validation
Passed on Claude Code 2.1.302View the source on GitHub Back to the list
What the scan recorded
- Footprint
- Unavailable or incomplete. Recorded events and calls below may omit access.
- Observed events
- subagent spawns, every prompt, /"(?:)"/ calls, Read calls, Agent calls, Workflow calls
- Hooks
turn.start {turnId=/"(?:)"/}, agent.spawn {tool_use_id=/"(?:)"/}, turn.complete {turnId=/"(?:)"/}, ui.render {component=AbovePrompt}, ui.render {component=SessionMode}, ui.render {component=Pane, requestId=?}, turn.start {turnId=/"(?:)"/}, agent.spawn {tool_use_id=/"(?:)"/}, session.start {cwd=/"(?:)"/}, session.measure {context has {window=/"(?:)"/}, command.run {command=dp}, prompt.submit {text=/"(?:)"/}, session.end {reason=/"(?:)"/}, session.start {cwd=/"(?:)"/}, turn.complete {turnId=/"(?:)"/}, tool.call {tool=/"(?:)"/}, turn.step {turnId=/"(?:)"/}, tool.call {tool=Read}, prompt.submit {text=/"(?:)"/}, session.append {door=/"^(?:prompt|queue)$"/}, tool.call {tool=/"(?:)"/}, turn.step {turnId=/"(?:)"/}, prompt.submit {text=/"(?:)"/}, tool.call {tool=Agent}, agent.spawn {tool_use_id=/"(?:)"/}, tool.call {tool=Workflow}, tool.describe {tool=Workflow}, tool.call {tool=Workflow}, agent.spawn {tool_use_id=/"(?:)"/}, session.start {cwd=/"(?:)"/}, session.start {cwd=/"(?:)"/}- API calls
$.agent.list, $.clock.after, $.clock.now, $.clock.sleep, $.command.list, $.command.register, $.env.get, $.fs.exists, $.fs.list, $.fs.read, $.fs.stat, $.http.fetch, $.model.complete, $.process.run, $.prompt.fill, $.session.cwd, $.session.id, $.session.messages, $.session.usage, $.settings.read, $.state.get, $.state.set, $.store.delete, $.store.get, $.store.keys, $.store.set, $.tool.register, $.ui.close, $.ui.invalidate, $.ui.log, $.ui.open, $.ui.panes, $.ui.resolve, $.ui.toast, $.ui.view- Validation on 2.1.302
- Passed
- Scanned revision
6f97fbf3fa1f- Marketplace passed
- alex-mods
- UI rewrite review
- UI hook source contains control-character strings. Review any values passed to next() as rewritten props.text; the scanner does not trace whether these strings reach that call. dispatch-pilot/hooks/decision/redact.ts:30, dispatch-pilot/hooks/decision/redact.ts:31, dispatch-pilot/hooks/decision/workflow-script.ts:516, dispatch-pilot/hooks/decision/workflow-script.ts:523, dispatch-pilot/hooks/decision/workflow.ts:469
These are the hooks and API calls reported by claude plugin validate. Access levels describe the widest reach of those calls and are not safety ratings. Validation is a static check, not a runtime test, and a passing result does not prove that a mod works or is safe. How the scan works.
More from alexcz-a11y/claude-mods
- prefer-search-toolsauto 模式(以及 bypass 模式)插给模型的那段提醒里,有一句让模型在 Bash 里用 grep 和 find 搜索。这个 mod 只拿掉这一句,换成用 Grep 工具搜内容、用 Glob 工具找文件名;提醒的…
Badge for your README
Authors can show the latest scan result in their README. Each badge links to this page.
[](https://mods.aidojo.si/mods/alexcz-a11y--claude-mods--dispatch-pilot/)
[](https://mods.aidojo.si/mods/alexcz-a11y--claude-mods--dispatch-pilot/)