ModsCode / Claude Code Mods / Ranking / task-orchestrator-mod
task-orchestrator-mod
Opt-in Claude Code mod for MCP Task Orchestrator: live work graph, gate band, and in-process orchestration hooks (early-access function-hooks API)
What it does
- Draws a band above the prompt, a pane, toasts
- Adds /to-band, /to-graph
- Watches tool calls and can refuse them
- Hooks agent.spawn, classic.PostToolUse, classic.Stop, classic.SubagentStop and 6 more
- Keeps its own state between turns
Good points
- No network calls
- Written in TypeScript
- Has a desktop-specific branch
- Handles errors
Code to start from
A starter ModsCode writes from the drawing it read, and the mod's own lines that draw, where there are any.
Starter (register.tsx)
import type { Register } from 'claude-code'
// Drawn like task-orchestrator-mod by Jeff Picklyk (pane, band, status, toast), as ModsCode read it from the mod's code:
// https://modscode.com/claude-mods/mods/jpicklyk-task-orchestrator-mod/
// Written by ModsCode from that reading, not copied from the mod: use it as you like.
// Each value below stands for one the mod works out at run time; put in your own.
export const register: Register = on => {
// the band above the prompt; what else goes there is drawn under it
on('ui.render', { component: 'AbovePrompt' }, async ($, e, next) => {
const { Box, Button } = $.ui.resolve(e)
const text = '…'
return (
<Box flexDirection="column">
<Box>
<Button label={`${text}`} plain />
<Button label="Hide" />
</Box>
{await next(e)}
</Box>
)
})
// the pane: open it with $.ui.open({ id: 'my-pane', title: "TO graph" })
on('ui.render', { component: 'Pane', requestId: 'my-pane' }, async ($, e) => {
const { Box, Text, Button } = $.ui.resolve(e)
const title = '…'
const title2 = '…'
const title3 = '…'
const header = '…'
return (
<Box flexDirection="column" paddingX={1}>
<Box flexDirection="row" flexWrap="wrap">
<Text dimColor>
{" › "}
</Text>
<Text bold>
{title}
</Text>
<Text dimColor>
{" › "}
</Text>
<Text bold>
{title2}
</Text>
<Text dimColor>
{" › "}
</Text>
<Text bold>
{title3}
</Text>
</Box>
<Box marginBottom={1}>
<Text>
{header}
</Text>
</Box>
<Box flexDirection="row" columnGap={1} marginBottom={1}>
<Button label="This feature" />
<Button label="Whole project" />
<Button label="Follow" />
<Button label="Hide done steps" />
<Button label="Reconnect" />
</Box>
</Box>
)
})
on('session.start', async ($, e, next) => {
const itemId = '…'
await $.ui.toast(`No active feature in this project.`)
await $.ui.toast(`Copied ${itemId}`)
return next(e)
})
}
The mod's own code
Shown under the mod's licence (MIT); its text is below. Keep the notice if you copy these lines.
claude-plugins/task-orchestrator-mod/src/band/index.ts lines 102–117 · Band
on('ui.render', { component: 'AbovePrompt' }, async ($, e, next) => {
if (e.props.hasSurvey || e.props.bodyColumns < MIN_BAND_COLUMNS) return next(e)
if (await read($, bandHidden)) return next(e)
// The smart line while the graph is scoped to an item, else today's in-flight line (model.ts).
const text = bandLine(await read($, graphSnapshot), e.props.bodyColumns)
if (text === undefined) return next(e)
const { Box, Button } = $.ui.resolve(e)
return h(
Box,
null,
h(Button, { key: 'band-open', label: text, plain: true, onPress: () => openPane($) }),
h(Button, { key: 'hide', label: 'Hide', onPress: () => update($, bandHidden, hide) }),
)
})claude-plugins/task-orchestrator-mod/src/graph-pane/index.ts lines 542–701 · Pane
on('ui.render', { component: 'Pane', requestId: PANE_ID }, async ($, e) => {
// Cross-session follow: (re)set the one-slot listener on every draw so a pane opened from the band or redrawn after a hot reload still follows (a draw after close is harmless: followActivity checks graphPaneOpen).
setRemoteAdvanceListener(itemId => void followActivity($, () => [itemId]))
const ui = $.ui.resolve(e)
const { Box, Text, Button, Svg } = ui
const snap = await read($, graphSnapshot)
const status = await read($, graphStatus)
const scope = await read($, graphScope)
const mode = await read($, graphScopeMode)
const detail = await read($, graphDetail)
const activity = await read($, graphActivity)
const showDone = await read($, graphShowDone)
const remote = (await read($, graphRemote)) ?? {}
const open = (id: string): Promise<void> => openDetail($, id)
const view: GraphView | null = snap
const switching = isSwitching(scope, snap)
const model = view === null ? null : cardsOf(view, activity.working, activity.changed, activity.blocked ?? {}, remote)
const steps = model === null ? null : stepsOf(model.cards)
const crit: CriticalPath = model === null || steps === null || view?.overview === true ? { ids: [], edges: new Set() } : criticalPath(model.cards, steps)
const lay = model === null || steps === null ? null : layoutTD(model.cards, steps, { bodyColumns: (e.props as { bodyColumns?: number }).bodyColumns, showDone: showDone || view?.overview === true, hasRoot: model.root !== undefined, wrap: view?.overview === true })
// Refresh is automatic (SSE or poll); Reconnect restarts the live source and shows only while degraded.
const controls = h(
Box,
{ key: 'controls', flexDirection: 'row', columnGap: 1, marginBottom: 1 },
// Two-state scope toggle; the active scope draws as the primary button.
h(Button, {
key: 'scope-feature',
label: 'This feature',
...(scope !== null ? { variant: 'primary' } : {}),
onPress: async () => {
const feature = await activeFeature($)
if (feature === null) $.ui.toast('No active feature in this project.')
else {
await update($, graphScopeMode, pinned)
await update($, graphScope, scopeTo(feature))
requestRefresh()
}
},
}),
h(Button, { key: 'scope-project', label: 'Whole project', ...(scope === null ? { variant: 'primary' } : {}), onPress: async () => {
await update($, graphScopeMode, pinned)
await update($, graphScope, scopeTo(null))
requestRefresh()
} }),
// Shown only while pinned: back to following this session's own TO activity (the bare /to-graph result).
...(mode === 'pinned'
? [h(Button, { key: 'scope-follow', label: 'Follow', onPress: async () => {
await update($, graphScopeMode, auto)
await update($, graphScope, scopeTo(await activeFeature($)))
requestRefresh()
} })]
: []),
...(lay !== null && view?.overview !== true && lay.doneSteps.length > 0
? [h(Button, { key: 'done-steps', label: showDone ? 'Hide done steps' : 'Show done steps', onPress: () => update($, graphShowDone, v => !v) })]
: []),
...(isDegraded(status) ? [h(Button, { key: 'reconnect', label: 'Reconnect', onPress: async () => {
await update($, graphReconnectRequest, bump)
requestReconnect(await read($, graphSubscribers))
} })] : []),
)
if (view === null || model === null || steps === null || lay === null) {
return h(Box, { flexDirection: 'column' }, controls, h(Text, { key: 'loading' }, status.lastError ?? 'Loading the work graph…'))
}
// A scope switch is still loading: say so, and draw nothing of the old scope (the header would claim it).
if (switching) return h(Box, { flexDirection: 'column' }, controls, h(Text, { key: 'switching' }, `Loading ${loadingTitle(scope, snap)}…`))
// With a breadcrumb the scope is named there; the header keeps the counts and the live source.
const trail = view.trail ?? []
const header = [...(trail.length > 0 ? [] : [scopeHeader(view)]), summaryLine(view), `live: ${status.liveSource}${status.refreshing ? ' · refreshing' : ''}`].join(' · ')
const crumbs = trail.flatMap((c, i) => {
const last = i === trail.length - 1
const sep = i > 0 ? [h(Text, { key: `crumb-sep:${c.id}`, dimColor: true }, ' › ')] : []
if (last) return [...sep, h(Text, { key: `crumb:${c.id}`, bold: true }, cut(c.title, 40))]
const target = c.id === view.rootId ? null : c.id
return [
...sep,
h(Button, {
key: `crumb:${c.id}`,
label: cut(c.title, 28),
plain: true,
onPress: async () => {
await update($, graphDetail, () => null)
await update($, graphScopeMode, pinned)
await update($, graphScope, scopeTo(target))
requestRefresh()
},
}),
]
})
const body: unknown[] = []
if (view.truncated) body.push(h(Text, { key: 'truncated', dimColor: true }, view.overview === true ? 'More children than shown.' : 'Large subtree: showing only the shallowest 150 items.'))
if (view.error !== undefined) body.push(h(Text, { key: 'error', color: 'red' }, view.error))
if (view.nodes.length === 0) {
body.push(h(Text, { key: 'empty' }, 'No items in this scope.'))
} else {
body.push(
h(
Box,
{ key: 'legend', flexDirection: 'row', flexWrap: 'wrap', rowGap: 1, marginBottom: 1 },
...legendItems().map(item => h(Box, { key: `legend-${item.key}`, backgroundColor: item.color, paddingX: 1, marginRight: 1 }, h(Text, { color: '#ffffff' }, item.text))),
h(Box, { key: 'legend-ready', backgroundColor: READY, paddingX: 1, marginRight: 1 }, h(Text, { color: '#ffffff' }, '○ ready')),
h(Text, { key: 'legend-edges', dimColor: true }, '┄ contains ╌ open blocker ─ satisfied'),
...(crit.ids.length > 0 ? [h(Text, { key: 'legend-crit', color: CRIT_COLOR }, ' ━ critical path')] : []),
...(view.nodes.some(n => n.planLabel !== undefined) ? [h(Text, { key: 'legend-label', dimColor: true }, ' Tn = plan label')] : []),
),
)
if (lay.tooWide !== null) {
body.push(h(Text, { key: 'too-wide', dimColor: true }, `Graph is ${lay.tooWide} columns wide; the pane shows ${lay.cols}. Widen the pane or open a smaller scope.`))
}
body.push(...canvasOf(ui, { model, lay, desktop: e.surface !== 'terminal' && Svg !== undefined, cell, steps: steps.max, open, count: view.nodes.length, overview: view.overview === true, extraChars: extrasChars(detail?.lines ?? null, trail.map(c => c.title)), crit }))
}
if (detail !== null) {
// The detail is its own bordered panel: title line bold, then facts, then a row of actions.
const facts = [
...detail.lines.map((line, i) => h(Text, { key: `detail-${i}`, ...(i === 0 ? { bold: true } : {}) }, line)),
h(Text, { key: 'detail-uuid', dimColor: true }, `uuid: ${detail.itemId}`),
]
const working = (activity.working[detail.itemId] ?? []).map(w =>
h(Text, { key: `working-${w.agentId}` }, `working: ${w.seat}${w.model !== undefined ? ` · ${w.model}` : ''} · ${w.agentId === 'main' ? 'main loop' : id8(w.agentId)}`),
)
const actions = [
h(Button, {
key: 'detail-copy',
label: 'Copy UUID',
onPress: async press => {
let copied = false
try {
const res = (await $.ui.copy({ text: detail.itemId, surface: press.surface })) as { isCopied?: boolean; value?: { isCopied?: boolean } }
copied = res.isCopied === true || res.value?.isCopied === true
} catch {
copied = false
}
// $.ui.copy has no path on a remote surface (the desktop app) yet: fall back to the OS clipboard tool.
if (!copied) copied = await hostCopy($, detail.itemId)
$.ui.toast(copied ? `Copied ${detail.itemId}` : `Copy unavailable here. UUID: ${detail.itemId}`)
},
}),
...(detail.itemId !== view.scopeId
? [
h(Button, {
key: 'detail-open-graph',
label: 'Open graph',
onPress: async () => {
await update($, graphDetail, () => null)
await update($, graphScopeMode, pinned)
await update($, graphScope, scopeTo(detail.itemId))
requestRefresh()
},
}),
]
: []),
h(Button, { key: 'detail-close', label: 'Close detail', onPress: () => update($, graphDetail, () => null) }),
]Licence text (MIT)
MIT License Copyright (c) 2025 Jeff Picklyk Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
This code is an example in the code reference: Band above the prompt
How to build this: Pane (panel), Deny a tool call
What its code calls
runs programsreads filesenv varsdrives Claudetool callsrememberscalls tools
What its code shows
Read from the source at commit 0d7d801d62bb on 2026-10-11. The first group decides whether a mod is listed; the rest is for you to weigh. A later commit is not read until the next crawl.
| Decides the listing: code you can check | |
|---|---|
| Entry read | Every module hooks/hooks.json names was read.claude-plugins/task-orchestrator-mod/hooks/register.ts |
| Parses | Every file parses as TypeScript or JavaScript.35 files parsed |
| All its code is in the mod | Every import is a file of the mod or the engine's own module; no packages, no require.all its code is in the mod |
| Readable | No minified lines, no eval or new Function, no import of a computed path, no encoded blobs.no minified code, no eval, no encoded blobs |
| Calls in the open | Network and programs are reached as $.http.fetch and $.process.run in place, never passed around.every call outside goes through $ in the open |
| For you to weigh before installing | |
| Where it draws | The terminal, the desktop app, both, or nothing (hooks only: a guard or a prompt rewriter), as its code says. Shown, never a reason to leave a mod out.draws in the terminal and the desktop app, the same in both |
| Install line | Whether its repository's own .claude-plugin/marketplace.json lists it, so it installs from that marketplace with the lines on its page.listed in its repository's .claude-plugin/marketplace.json: installs from there |
| Programs on Windows | Whether every program it starts through $.process exists on Windows (tail, date, open, osascript, /usr/… do not).runs no program a Windows PC lacks |
| Network | Whether every fetch names its https host in the code.no network |
| Programs | Whether every program it starts is named in the code and is not a shell, an interpreter or a network tool.claude-plugins/task-orchestrator-mod/src/graph-data/index.ts:148 a program named at run time; claude-plugins/task-orchestrator-mod/src/graph-pane/index.ts:76 a program named at run time |
| Reads and sends | Whether it both reads something of yours (files, environment variables, the transcript, settings) and reaches outside (the network, an MCP server, another session, telemetry).reads environment variables, files and reaches an MCP server |
| Keys in the source | Whether the source holds what looks like an API key or a private key.no keys in the source |
| Licence | An open licence (MIT, Apache-2.0, BSD, ISC, Unlicense, 0BSD, CC0) with its text lets ModsCode show its code; otherwise the code is linked.MIT |
| Settings hooks | Hooks declared in hooks/hooks.json or plugin.json beside the module: a command (a shell when it has no args), an HTTP request, an MCP tool or a model prompt at an event. They run outside $.no settings hooks: its function hooks alone |
| Servers it ships | MCP servers (a program, or a URL; a headersHelper is a shell command; a .mcpb bundle is unpacked at run time), LSP servers, monitors (a shell command for the whole session) and bin/ (on the PATH of the Bash tool).ships no MCP or LSP server, no monitor, no bin/ |
| Skills, commands, agents | What its Markdown declares beyond text: allowed-tools granted for the turn, hooks added when a skill is invoked, a command run before Claude reads the skill.no skills, commands, agents or workflows |
| Tools and MCP | Whether every tool it calls through $.tool.call and $.mcp.call is named in the code, and none is a shell (Bash through the tool runs a shell without $.process).claude-plugins/task-orchestrator-mod/src/graph-data/index.ts:142 an MCP tool named at run time; claude-plugins/task-orchestrator-mod/src/graph-pane/index.ts:95 an MCP tool named at run time; claude-plugins/task-orchestrator-mod/src/graph-pane/index.ts:125 an MCP tool named at run time; claude-plugins/task-orchestrator-mod/src/graph-pane/index.ts:163 an MCP tool named at run time; claude-plugins/task-orchestrator-mod/src/phase-guard/index.ts:63 an MCP tool named at run time |
| Asks of the model | Completions it runs ($.model.complete, fork, classify), subagents it spawns, prompts it submits as you, and the tools, commands and agents it adds.adds command to-band, adds command to-graph, agent.spawn |
| Events it hooks | By what a hook there can do: gates (refuse a tool call, a prompt, a setting, another plugin), shapes (what the model reads: the system prompt, tool descriptions, each request), reads (your input, the events of the settings hooks, every event with *), and the rest.gates: prompt.submit, command.run (to-band, to-graph), tool.call, classic.SubagentStop, classic.Stop · reads: classic.PostToolUse · also: state.set, turn.complete, ui.render (AbovePrompt, Pane), session.start (+3) · timers: clock.after, clock.every |
| Reads, by name | The environment variables and files it names, and whether it reads the transcript, settings.json, the prompt box or the mouse selection.env: TASK_ORCHESTRATOR_API_URL, TASK_ORCHESTRATOR_API_TOKEN, TASK_ORCHESTRATOR_HOME, HOME, USERPROFILE (+3); files: a path computed at run time, …/… |
| Writes, by name | The files and environment variables it names, and whether it changes settings, the prompt box, the clipboard, the conversation (a row the model reads) or another session.the clipboard |
This is a reading of the code text, not a security review: it does not run the mod, and it cannot see what a program the mod starts does. The check before installing is yours. How mods are chosen
Install it
Check it yourself before you install it. Mods run unsandboxed, with your permissions, and ModsCode does not vouch for what this code does. Read its code (or run claude plugin validate . in its folder), weigh what it reaches, and install the commit you read.
From its marketplace
Its repository is its marketplace. Add that marketplace (in the terminal with these commands, or in the desktop app's plugin settings) and install the mod from it:
/plugin marketplace add jpicklyk/task-orchestrator/plugin install task-orchestrator-mod@task-orchestrator-marketplace/reload-pluginsThis installs the marketplace's latest commit, which may not be the one read here (0d7d801d62bb). Compare them before you install. /reload-plugins loads it into a session that is already open.
Try the commit ModsCode read
No marketplace needed. In a terminal, clone the repository, check out the commit read, validate the mod, and load it for one session:
git clone https://github.com/jpicklyk/task-orchestratorgit -C task-orchestrator checkout 0d7d801d62bb82c4a201ee61f1787194c8462ae6claude plugin validate task-orchestrator/claude-plugins/task-orchestrator-modclaude --plugin-dir task-orchestrator/claude-plugins/task-orchestrator-modThe validate line lists the events the mod hooks and what it asks Claude Code to do, without running it. The last line loads the mod for this session only and writes nothing to your settings. If the author has rewritten history, the commit may no longer exist.
Ask Claude about it
With the ModsCode connector added, Claude can read this mod's code and what it reaches, and build on them. Ask it like this:
Read the ModsCode mod "jpicklyk-task-orchestrator-mod" with its code and make me one like itSearch them from Claude: the MCP connector
Similar mods
- code-modernizationAnthropic★ 37,653
- crit-tuiTomasz Tomczyk★ 1,196
- remctlFederico Viticci★ 595
- localvoxtral-modlocalvoxtral★ 59
- localvoxtral-remotelocalvoxtral★ 59
- corgi-bandAndrii Klymiuk★ 37
Badge for your README
If this is your mod, paste this into your README. It links to this page.
[](https://modscode.com/claude-mods/mods/jpicklyk-task-orchestrator-mod/)