agentleFS
tuana-vntuana-vn/system-reverse-engineer/.github/skills/wbs-generation/SKILL.md

wbs-generation

Skill0 starsChanged 2 months agoMITView on GitHub

Generate a traceable implementation Work Breakdown Structure from verified detailed-design artifacts, current source anchors, requirement/TDD traceability, dependencies, and verification obligations. Use only after detailed designs are verified.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 9 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not jump from TDD prose directly to generic tasks.

    line 13
  2. Distinctive:

    test design if available

    line 35
  3. Distinctive:

    Do not add "nice to have" refactoring.

    line 55
  4. Distinctive:

    If a WBS task's primary deliverable is a subordinate type, one of these must be true:

    line 96
  5. Distinctive:

    Do not map a task titled for component/type A to an artifact whose registered responsibility is component/type B unless the artifact …

    line 99
  6. Very specific:

    If semantic ownership cannot be proven, stop with WBS_BLOCKED_DESIGN_GAP.

    line 101
  7. Distinctive:

    Before emitting the WBS, create an internal/exposed consistency table:

    line 104
  8. Distinctive:

    Do not invent ordering solely from file order.

    line 142
  9. Distinctive:

    Every code-changing task needs a verification path.

    line 146
  10. Distinctive:

    No verified design artifact may be orphaned.

    line 162
  11. Very specific:

    Do NOT estimate hours/days/story points in this skill.

    line 189

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Implementation WBS Generation

Purpose

WBS is WHAT must be implemented, not an effort estimate.

Do not jump from TDD prose directly to generic tasks.

Inputs must include VERIFIED detailed designs.

Requirement
→ Gap
→ TDD
→ Design Artifact
→ VERIFIED Detailed Design
→ WBS Task
→ Verification

Inputs

  • authoritative requirement
  • verified gap analysis
  • verified TDD
  • design artifact registry
  • all DD_VERIFIED detailed designs
  • detailed-design verification reports
  • test design if available
  • current repository

1. Task Derivation

Derive tasks only from verified design decisions.

Task categories:

  • production code
  • API/contract
  • integration
  • data/persistence
  • configuration
  • migration/compatibility
  • tests
  • build/deployment
  • documentation
  • cleanup only if required by approved design

Do not add "nice to have" refactoring.

2. Task Granularity

A task should be small enough to have:

  • one clear outcome
  • concrete source/design anchors
  • explicit dependencies
  • verifiable done criteria

But do not split mechanical edits into dozens of meaningless microtasks.

3. Stable IDs

Use:

WBS-001, WBS-002, ...

3A. WBS ↔ Design Semantic Identity

Traceability is not satisfied merely because a WBS row contains an existing Artifact ID.

For every WBS task, identify its primary implementation deliverable and prove that the referenced design artifact(s) own that deliverable.

Required semantic chain:

WBS task title / objective / primary deliverable
↔ registry artifact name + responsibility
↔ VERIFIED DD owned implementation unit(s)
↔ TDD design decision

Rules:

  1. An Artifact ID reference is necessary but not sufficient.
  2. A type/class/interface being merely mentioned, imported, referenced, or linked from a DD does not prove that the artifact owns that implementation unit.
  3. If a WBS task's primary deliverable is a subordinate type, one of these must be true:
    • the registry/DD explicitly declares that subordinate type inside the owning artifact scope, or
    • the design artifact decomposition must be revised to track it separately.
  4. Do not map a task titled for component/type A to an artifact whose registered responsibility is component/type B unless the artifact explicitly owns A.
  5. If semantic ownership cannot be proven, stop with WBS_BLOCKED_DESIGN_GAP; do not manufacture a mapping from nearby TDD prose.

Before emitting the WBS, create an internal/exposed consistency table:

WBS IDPrimary DeliverableReferenced Artifact IDsRegistry ResponsibilityDD-Owned UnitSemantic Match?Evidence

Required threshold:

WBS ↔ artifact semantic identity = 100%

4. WBS Table

WBS IDDesign Artifact IDsTaskExisting Source Anchor / Proposed LocationChange TypeDependenciesRequirement IDsTDD/DD ReferencesVerificationDone Criteria

Change Type:

  • ADD
  • MODIFY
  • REMOVE
  • CONFIGURE
  • MIGRATE
  • TEST
  • DOCUMENT
  • VERIFY

5. Dependency DAG

Derive implementation ordering from artifact dependencies and code/runtime dependencies.

Create:

WBS IDDepends OnDependency ReasonHard/Soft

Detect circular dependency.

Do not invent ordering solely from file order.

6. Verification Mapping

Every code-changing task needs a verification path.

Map to:

  • existing test to update
  • new test ID/viewpoint
  • source/build verification
  • runtime/integration verification
  • manual verification only when automation is not feasible

7. Traceability

Create:

Requirement IDGap IDTDD Decision IDArtifact IDDDWBS IDsTest/Verification IDsCoverage

No verified design artifact may be orphaned.

8. WBS Gate

Before WBS_READY_FOR_VERIFICATION:

[ ] all DD_VERIFIED artifacts have implementation/verification task coverage
[ ] every WBS task maps back to verified design
[ ] every WBS task primary deliverable semantically matches the responsibility/owned unit of its referenced artifact(s)
[ ] no task relies on a merely-mentioned DD/TDD symbol as proof of artifact ownership
[ ] source anchors/proposed locations are concrete
[ ] dependencies are explicit
[ ] code-changing tasks have done criteria and verification
[ ] config/migration/docs/test work is included when required
[ ] no unapproved architecture/refactoring was added
[ ] no HIGH/CRITICAL source gap is hidden inside a task

Statuses:

  • WBS_READY_FOR_VERIFICATION
  • WBS_BLOCKED_DESIGN_GAP
  • WBS_BLOCKED_SOURCE_GAP

Important

Do NOT estimate hours/days/story points in this skill. Effort estimation is a separate concern.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.