- Home
- Skills
- Agents & Orchestration
- planning and task breakdown
planning and task breakdown
Decompose complex requirements into small, verifiable vertical slices with explicit dependency mapping and checkpoints.
Free
Works with the AI Agents you already use
planning and task breakdown
Example session with this skill installed
I need to add a user profile page where users can update their bio and avatar. Here is the existing User schema and API structure. Break this down into an implementation plan.
- Read your context and instructions
- Compiled the planning and task
I have analyzed the requirements for the user profile feature. I will now create the implementation plan and task list.
Implementation Plan: User Profile Management
Overview
Enable users to manage their public-facing identity by updating their bio (text) and avatar (image URL). This involves extending the API for partial updates and creating a dedicated settings UI.
Architecture Decisions
- Use HTTP PATCH for profile updates to allow partial bio/avatar changes.
- Implement client-side image validation for avatar URLs before submission.
Task List
Task 1: API Update for Profile Fields
- Description: Add bio and avatar fields to the user update endpoint logic.
- Acceptance criteria:
- API accepts
PATCH /api/userwith bio and avatar strings. - Database updates correctly for both fields.
- API accepts
- Verification:
- Tests pass:
npm test api/user - Manual check: Verify DB entry via CLI after request.
- Tests pass:
- Dependencies: None
- Estimated scope: Small (2 files)
Task 2: Profile Settings UI Component
- Description: Build the form UI for bio text and avatar URL input.
- Acceptance criteria:
- Form displays current user data on load.
- Submit button is disabled if no changes are made.
- Verification:
- Build succeeds:
npm run build - Manual check: Render component in storybook/dev.
- Build succeeds:
- Dependencies: Task 1
- Estimated scope: Medium (3 files)
Task 3: Integration and Toast Notifications
- Description: Wire the UI to the API and add success/error feedback.
- Acceptance criteria:
- Successful save shows a "Profile updated" toast.
- Network errors display a specific error message.
- Verification:
- Tests pass:
npm test e2e/profile
- Tests pass:
- Dependencies: Task 2
- Estimated scope: Small (2 files)
Checkpoint: User Profile Flow
- All tests pass
- Profile updates persist across page reloads
- Review UI alignment with design system
Next steps
- Review
tasks/plan.mdfor architectural alignment. - Initialize Task 1 in the development environment.
- Confirm image storage strategy if avatar requirements change to file uploads.
Connects securely to your tools. The creator never sees your data.
About this skill
Complex features often stall when agents or developers tackle too much at once, leading to tangled code and missed edge cases. This skill enforces a systematic decomposition process, turning vague requirements into a strictly ordered graph of small, verifiable tasks.
What it does
- Architectural mapping identifies existing patterns, dependencies, and risks before a single line of code is written.
- Vertical slicing ensures each task delivers a functional, testable path through the stack rather than isolated horizontal layers.
- Dependency sequencing generates a bottom-up implementation order to ensure foundations are built before dependents.
- Task sizing breaks work into XS-M units (under 5 files) to maximize agent reliability and focus.
- Automated documentation generates
tasks/plan.mdandtasks/todo.mdto maintain state across session boundaries.
How it works
- Enter Plan Mode to analyze the spec and codebase in a read-only state to identify patterns and risks.
- Map Dependencies to visualize the relationship between schema, APIs, and UI components.
- Execute Vertical Slicing to group related changes into functional units rather than technical layers.
- Generate Task List with explicit acceptance criteria, verification steps, and file estimates.
- Establish Checkpoints to mandate build and test verification after every 2-3 tasks.
Frameworks & tools
This skill is framework-agnostic and works with any codebase using markdown-based planning or external trackers like GitHub Issues, Jira, and Linear.
Why this beats prompting it yourself
Standard prompts often result in "hallucinated" progress where an agent claims a feature is done despite failing builds. This skill forces a dependency-aware structure and explicit verification steps that survive context window resets.
Use cases
- Decomposing a high-level PRD into a multi-day implementation roadmap.
- Splitting a large monolithic task that has become too complex for a single agent session.
- Planning parallel work streams for multiple agents to work on independent feature slices.
- Generating a structured task list for an external tracker like Linear or GitHub.
Known limitations
This skill is not for single-file hotfixes or minor config changes where the scope is obvious. It requires an existing spec or clear set of requirements to be effective.
Requires an existing spec or requirements. Best for multi-file features, not trivial hotfixes. Task tracking defaults to markdown unless an external tracker is specified.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
2 uses
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account