plan-tasks
Skill31 starsChanged 6 months agoMITView on GitHub
Creates a structured implementation task breakdown for a structured change, synthesizing proposal, design, specs, and any supplied test-plan obligations into self-contained per-task files; use when the tasks artifact is the next step in a change.
We split this skill into the rules it gives an agent
8 rules. 0 are common, 2 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 7
Create a small set of self-contained delivery tasks from the proposal, specifications, design, any supplied test plan, and relevant …
- Very specific: line 12
Read the proposal, design, specifications, and relevant repository context.
- Very specific: line 44
When a test plan is supplied, assign each INT-* and E2E-* obligation to the task that first makes its boundary or journey executable.
- Very specific: line 46
Do not assign acceptance exploration or HT-* execution to implementors.
- Very specific: line 56
Write tasks under <change-dir>/tasks/.
- Very specific: line 87
When a test plan is supplied, include - test-plan.md for <assigned INT/E2E obligations> under You MUST read and the ## Test Plan section …
- Very specific: line 109
When a test plan is supplied, add a ## Test Plan section with the assigned INT-* and E2E-* obligations and relevant strategy.
- Distinctive: line 113
Keep all variants of one behavior together.
“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Plan Tasks
Create a small set of self-contained delivery tasks from the proposal, specifications, design, any supplied test plan, and relevant repository context. Each task goes to a skilled agent with no prior conversation; include necessary decisions, paths, constraints, scenarios, and automated-test obligations without dictating line-by-line implementation.
Read the proposal, design, specifications, and relevant repository context. When test-plan.md exists
or the caller supplies a test plan, read it and include its obligations; otherwise plan from the
available definition artifacts without requiring or inventing test-plan content.
Size before decomposing
Estimate the whole change from novelty, uncertainty, architectural reach, migration risk, integration breadth, and verification burden. Do not equate file, requirement, scenario, provider, or UI counts with task count.
| LOE | Typical tasks |
| Small | 1 |
| Medium | 2 |
| Large | 3–4, preferably 3 |
| XL | 5+, starting at 5 |
Choose the smallest fitting band. Mechanical propagation across layers or adapters usually remains one task. For Medium-or-larger changes, or unclear provider, migration, cutover, feasibility, or cleanup boundaries, read references/task-sizing.md.
Form delivery units
A task is an independently meaningful outcome a generalist can implement, test, and review in one focused session. Prefer vertical slices that include their schema, config, logic, interfaces, tests, and documentation.
Split only for a distinct outcome, implementation risk, real dependency or release gate, or substantial independently verifiable foundation. Merge boundaries that create broad unfinished plumbing, duplicate edits, or a large implicit handoff. Fold ordinary scaffolding, migrations, refactors, test setup, and documentation into the outcome they support.
When a test plan is supplied, assign each INT-* and E2E-* obligation to the task that first makes
its boundary or journey executable; a cross-task E2E belongs to the final task completing that journey.
Do not assign acceptance exploration or HT-* execution to implementors. Unit cases remain implementation-time
decisions.
Compare the result with the independent LOE estimate. Merge coupled candidates above the band; never invent work to reach a count. Use repository context to resolve ordinary planning choices. Stop only when approved source artifacts contradict each other enough that safe implementation cannot be planned. Do not ask the user to approve the task count or grouping.
Write the tasks
Write tasks under <change-dir>/tasks/. Use an unnumbered <slug>.md for one task or ordered
<NN>-<slug>.md files for multiple tasks. Use real repository paths and no placeholders.
Copy every name the design or specs fix verbatim: test hooks and attributes, selectors, routes, file paths, commands, config and schema keys, and public identifiers. Other tooling may depend on the exact spelling, so do not rename, namespace, or offer alternatives or a choice of scheme. If a fixed name looks wrong, keep it and record the concern as an open assumption in the task's Background and in your report.
For one task, exact references to the design and specs may replace copied artifact text. When a test
plan is supplied, also cite it for the assigned INT-* and E2E-* obligations:
# Task: <Title>
## Goal
<Outcome and reason>
## Background
You MUST read:
- `design.md` for <relevant decisions>
- `specs/<capability>/spec.md` for <requirements and scenarios>
<Relevant paths, constraints, and context>
## Test Plan
- `<INT-* or E2E-*>`: <strategy and completion signal>
## Done When
<Concrete delivery signals, including the assigned automated obligations>When a test plan is supplied, include - test-plan.md for <assigned INT/E2E obligations> under
You MUST read and the ## Test Plan section shown above. Otherwise omit both and use ordinary
completion signals under ## Done When.
For multiple tasks, each file must stand alone and must not refer to another task:
# Task: <Title>
## Goal
<Outcome and reason>
## Background
<Motivation, design decisions, exact paths/APIs, and constraints>
## Spec
<Relevant requirements and scenarios copied verbatim>
## Done When
<Scenarios and automated obligations pass, plus concrete delivery signals>When a test plan is supplied, add a ## Test Plan section with the assigned INT-* and E2E-*
obligations and relevant strategy; otherwise omit that section and test-plan-specific completion
signals.
Keep all variants of one behavior together. If tasks genuinely share a scenario, copy it into each and state that task's portion. A standalone refactor is justified only when it is independently risky and reviewable; otherwise keep refactoring with the behavior it enables.
Order multiple tasks by real dependencies and use matching zero-padded prefixes. Write
<change-dir>/tasks.md with one linked checkbox per task:
- [ ] [<Task title>](tasks/01-<slug>.md)
- [ ] [<Task title>](tasks/02-<slug>.md)The task title MUST be an actual Markdown link. A path in an inline code span is not a link.
Report the task count, titles, and index path. Do not invoke another lifecycle skill.
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- oopsla5xx/oops-skills★ 4same name
- zahardev/aicontext★ 2same name
- jayden-dang/skills★ 1same name
- davidscotttufts/hyvemind★ 0same name
- Zaheer-verse/deliberate-dev-claude★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository30 other files it gives its agents.
- AGENTS.mdAGENTS.md
- docs/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- releaseSkill
- releaseSkill
- ask-questionsSkill
- call-agentSkill
- designSkill
- finalize-prSkill
- fix-prSkill
- handoffSkill
- implement-changeSkill
- initSkill
- orchestrate-changeSkill
- prepare-acceptanceSkill
- proposal-reviewSkill
- proposeSkill
- push-prSkill
- review-approachSkill
- review-assumptionsSkill
- review-specSkill
- review-tasksSkill
- session-reportSkill
- simple-planSkill
- specSkill
- task-complianceSkill
- test-flowsSkill
- test-planSkill
- verify-behaviorSkill
- wait-ciSkill
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.

