integrate-todo-list
Skill6 starsChanged 9 days agoApache-2.0View on GitHub
TodoWrite panel for an already-approved in-app useAtlasChat UI. New AI features use integrate-fusion-agent. Triggers: TodoWrite, todo panel — only when useAtlasChat already exists.
In this repository34 other files it gives its agents.
- aura-reviewSkill
- call-cognite-functionSkill
- code-qualitySkill
- connect-atlas-mcpSkill
- correctness-and-error-handlingSkill
- create-agentSkill
- create-client-toolSkill
- dependencies-auditSkill
- designSkill
- dm-graph-traversalSkill
- dm-limits-and-best-practicesSkill
- flows-app-briefSkill
- flows-code-reviewSkill
- flows-design-reviewSkill
- flows-external-app-submitSkill
- flows-review-checksSkill
- graph-viewerSkill
- handle-startup-stateSkill
- hide-show-shellSkill
- integrate-atlas-chatSkill
- integrate-file-viewerSkill
- integrate-fusion-agentSkill
- migrate-app-to-flowsSkill
- performanceSkill
- pull-changes-resolve-conflictsSkill
- reveal-3dSkill
- securitySkill
- setup-flows-authSkill
- setup-python-toolsSkill
- skill-creatorSkill
- start-flows-appSkill
- test-coverageSkill
- tygenSkill
- use-topbarSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 2 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 9
Add a TodoWrite tool and panel so the in-app Atlas agent can show task progress.
- Distinctive: line 17
Before writing anything, read:
- Very specific: line 30
Read each file from <skill-dir>/code/ and write it into src/todo/ with the same filename:
- Very specific: line 43
All files use relative imports (./types, ./TodoContext, etc.) — no changes needed.
- Very specific: line 145
Run the app's type-check command (typically pnpm tsc --noEmit) and confirm there are no errors.
- Distinctive: line 146
If the project has tests, run them to confirm nothing regressed.
- Very specific: line 154
Update task status in real-time (pending → in_progress → completed)
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Integrate Todo List
Add a TodoWrite tool and panel so the in-app Atlas agent can show task progress.
Skip if the app uses the Atlas / EOS sidebar (integrate-fusion-agent). Prerequisite: useAtlasChat + src/atlas-agent/ + @sinclair/typebox from integrate-atlas-chat.
Step 1 — Read the app
Before writing anything, read:
package.json— confirm@tabler/icons-reactis installed; if not, install it with the app's package managersrc/App.tsx— find where to addTodoProvider- The file that calls
useAtlasChat(likelysrc/chat/useChatViewModel.tsorsrc/App.tsx) — this is where the tool gets wired - The chat view component that renders messages — this is where
TodoPanelandTodoToolResultCardgo
Step 2 — Create the src/todo/ module
Find the skill directory by running find . -path "*/.agents/skills/integrate-todo-list/code" -type d from the project root.
Read each file from <skill-dir>/code/ and write it into src/todo/ with the same filename:
| File | Purpose |
types.ts | TodoItem and TodoList types |
TodoContext.tsx | React context + TodoProvider |
useTodoList.ts | Hook to read/write the todo list |
todoWriteTool.ts | createTodoWriteTool factory — AtlasTool with full CDF task-decomposition guidance |
useTodoWriteTool.ts | Hook that memoizes the tool with current state access |
TodoPanel.tsx | Card UI: progress bar + task rows |
TodoItemRow.tsx | Single row with animated status icons |
TodoToolResultCard.tsx | Compact summary card for tool call display |
All files use relative imports (./types, ./TodoContext, etc.) — no changes needed.
Step 3 — Wrap the app in TodoProvider
In src/App.tsx (or the root component), wrap the existing tree with <TodoProvider>:
import { TodoProvider } from './todo/TodoContext'; // adjust path to match app conventions
function App() {
return (
<TodoProvider>
{/* existing children */}
</TodoProvider>
);
}Step 4 — Wire the tool into useAtlasChat
In the file that calls useAtlasChat, add the following. Adjust import paths to match the app's conventions.
import { useRef, useCallback } from 'react';
import { useTodoList } from './todo/useTodoList';
import { useTodoWriteTool } from './todo/useTodoWriteTool';
// Inside the hook/component:
const { todos, setTodos } = useTodoList();
const todoWriteTool = useTodoWriteTool();
// Keep a ref so getAppContext always reads fresh state without re-creating the callback.
const todosRef = useRef(todos);
todosRef.current = todos;
const getAppContext = useCallback(() => {
const t = todosRef.current;
if (t.length === 0) return undefined;
const lines = t.map((item, i) => `${i + 1}. [${item.status}] ${item.content}`);
return `Current todo list:\n${lines.join('\n')}`;
}, []);
// Add to useAtlasChat options:
const { messages, send, isStreaming, progress, error, reset, abort } = useAtlasChat({
client: isLoading ? null : sdk,
agentExternalId: AGENT_EXTERNAL_ID,
tools: [todoWriteTool], // add alongside any existing tools
getAppContext,
});
// In the reset handler, clear the todo list:
const handleReset = useCallback(() => {
reset();
setTodos([]);
}, [reset, setTodos]);
// Expose todos in the return value so the view can render TodoPanel:
return { ..., todos };Step 5 — Render TodoPanel in the chat view
In the component that renders the chat input area, add <TodoPanel> above the input field:
import { TodoPanel } from './todo/TodoPanel'; // adjust path
// In the render:
<TodoPanel todos={todos} />
<YourChatInput ... />TodoPanel returns null when the list is empty, so it's safe to always render it.
Step 6 — Render TodoToolResultCard for tool call steps
In the component that renders per-message tool calls (typically a steps accordion or similar), branch on the tool name:
import { TodoToolResultCard } from './todo/TodoToolResultCard'; // adjust path
{toolCalls.map((tc, i) =>
tc.name === 'TodoWrite' ? (
<TodoToolResultCard key={i} toolCall={tc} />
) : (
<YourDefaultToolCallCard key={i} toolCall={tc} />
)
)}Step 7 — Verify
Run the app's type-check command (typically pnpm tsc --noEmit) and confirm there are no errors.
If the project has tests, run them to confirm nothing regressed.
Done
The agent can now use TodoWrite to create and track tasks. It will:
- Show a task panel as soon as it starts multi-step work
- Update task status in real-time (
pending→in_progress→completed) - Clear the list automatically when all tasks are done
- Inject the current task list into each prompt via
getAppContextso it knows where it left off
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

