codyswanngt/lisa-task-decomposition

Task Decomposition

Break work into ordered, well-scoped tasks that can be independently implemented and verified.

Decomposition Process

1. Identify Units of Work

• Break the work into the smallest units that are independently valuable
• Each unit should produce a verifiable outcome (a passing test, a working endpoint, observable behavior)
• Avoid tasks that are too large to complete in a single session
• Avoid tasks that are too small to be meaningful (e.g., "add an import statement")

1.5. Preserve Cross-Repo Containers, Split Leaf Work by Repository

Start from the right shape:

• A source PRD may span multiple repositories.
• Coordination containers may also span multiple repositories.
• Buildable leaf work units must be implementable inside exactly one repository.

That last point is a hard invariant — downstream validators (jira-validate-ticket, github-validate-issue, linear-validate-issue) gate writes on it, so a cross-repo leaf will fail to be created.

Apply this rule by layer:

| Layer | Repo scope | |-------|------------| | PRD / source initiative | MAY span repos — it describes the full initiative | | Epic, Story, Spike | MAY span repos — these are coordination containers | | Task, Bug, Sub-task, Improvement | MUST name exactly one repo — these are buildable leaf work units |

If a candidate work unit naturally touches multiple repos (e.g., "add field to backend API and consume it in mobile app"), do not write it as one ticket. Instead:

1. Split it into one work unit per repo (e.g., [backend-api] Add field to /users endpoint, [mobile-app] Display new field on profile screen).
2. Group the per-repo units under a single parent Story (or Epic, if the parent Story already exists). The parent stays cross-repo; the children do not.
3. Encode the order via Dependencies in step 4 — typically the producing repo (backend) blocks the consuming repo (frontend/mobile).
4. Tag each work unit with [repo-name] as a summary prefix so the repo is visible in tracker lists at a glance.

Reject any work unit whose acceptance criteria reference behavior in a different repo from the one it's scoped to. If you find yourself writing "and the frontend should also...", that's a signal to split.

This is the decomposition-time strategy (greenfield — you are creating the tickets now, so a cross-repo PRD can stay whole, its Epic/Story/Spike containers can stay cross-repo, and a parent Story + per-repo children is the natural shape). It is distinct from the work-time strategy in the repo-scope-split rule, which applies when an agent picks up an already-existing leaf ticket to implement and discovers it spans repos: there it narrows the original in place and spins off sibling work units rather than introducing a new parent. Use the phase-appropriate one; do not mix them.

2. Define Acceptance Criteria

For each task, define what "done" looks like:

• Be specific and measurable -- avoid vague criteria like "works correctly"
• Include both positive cases (what should work) and negative cases (what should be rejected)
• Reference exact behavior: error messages, status codes, output format, performance thresholds
• If a task modifies existing behavior, state both the before and after

3. Assign Verification Type

Each task must have a verification method. Choose the most appropriate:

| Verification Type | When to Use | |-------------------|-------------| | Unit test | Pure logic, data transformations, utility functions | | Integration test | Cross-module interactions, database operations, API contracts | | E2E test | User-facing workflows, multi-service interactions | | Manual verification | UI/UX behavior, visual correctness, one-time infrastructure changes | | Build verification | Compilation, type checking, linting, bundle size | | Deploy verification | Service health checks, smoke tests, monitoring dashboards |

4. Map Dependencies

• Identify which tasks must complete before others can start
• Order tasks so that each builds on a stable foundation
• Prefer independent tasks that can run in parallel where possible
• Flag external dependencies (other teams, services, permissions, data) that may block progress

4.5. Classify Any New Material Dependency

Step 4 maps dependencies between tasks. This step covers third-party dependencies a task proposes to add.

A dependency is material if its failure, disappearance, or bad update would break something a user can see, or would cost real time to replace. If a work unit proposes adding one, the work unit must, before it is buildable:

1. Name its trust class — exactly one of: mature ecosystem primitive, fast-moving standard implementation, build/development tool, runtime-critical service client, thin wrapper suitable for in-house ownership, or temporary/experimental dependency. See the dependency-trust-classes rule for what each class means and why it is trusted.
2. State the class's required evidence — the detection evidence that class demands, and whether product/human ratification is required before the work starts. Runtime-critical service clients and expiry extensions on temporary dependencies always require ratification; a work unit that needs it and does not say so is not ready.
3. Include updating .lisa/DEPENDENCY_DECISIONS.md in its acceptance criteria, so the record entry lands in the same change as the dependency.

If nobody can pick a class, that is the finding — resolve it before accepting the work unit, not after the package is installed.

4.6. Inherit the Confidence-Rebuild Kit When Ownership Moves In-House

Step 4.5 covers dependencies a task proposes to add. This step covers dependencies a task proposes to remove, replace, or internalize — vendor, fork-and-maintain, or reimplement ourselves.

When ownership moves in-house, the risk stops being "do we trust upstream" and becomes "did we prove we rebuilt the capability." A work unit that moves a material dependency in-house inherits all seven acceptance criteria of the dependency-internalization-kit rule, each written as an acceptance criterion of the work unit:

1. Real corpus — did we test it on real inputs, not toy examples?
2. Conformance fixtures — does the new code do what the dependency did?
3. Negative fixtures — does it still reject what it should reject?
4. Coverage as a gap detector — what behavior is still untested?
5. Provenance and license review — where did this code come from, and are we allowed to use it?
6. Migration and update plan — how do existing call sites move, and how does the new code stay current?
7. Rollback or replacement criteria — what would make us go back, and to what?

The only way out is an explicit non-material justification written into the work unit — one reviewable sentence saying why this dependency's failure or disappearance breaks nothing a user can see and costs no real time to replace. Silence is not a justification.

Do not over-apply the kit. A routine version bump of a trusted dependency within its existing trust class does not move ownership, so it does not inherit the kit — its bar is that trust class's own detection evidence and cadence. The same is true of swapping one third-party dependency for another, which moves trust to a different upstream rather than in-house. Add the kit only when ownership moves in-house; the exception is a bump taken as a fork or declined in favor of owning the code, which is an internalization regardless of how the ticket is titled.

5. Determine Execution Order

• Place foundational tasks first (types, schemas, interfaces, shared utilities)
• Follow with implementation tasks (business logic, handlers, services)
• Then integration tasks (wiring, configuration, API routes)
• Finish with verification tasks (test suites, documentation, cleanup)

6. Assign Required Skills

Map each task to the skills needed to complete it. This enables delegation to specialized agents or helps identify what expertise is required.

Output Format

## Task Breakdown

### Task 1: [[repo-name] imperative description]
- **Repository:** [single repo name, or N/A for Epic/Story/Spike]
- **Acceptance criteria:**
  - [specific, measurable criterion]
  - [specific, measurable criterion]
- **Verification:** [type] -- [how to verify]
- **Dependencies:** [none | task IDs that must complete first]
- **Skills:** [list of skills needed]

### Task 2: [[repo-name] imperative description]
- **Repository:** [single repo name, or N/A for Epic/Story/Spike]
- **Acceptance criteria:**
  - [specific, measurable criterion]
- **Verification:** [type] -- [how to verify]
- **Dependencies:** [Task 1]
- **Skills:** [list of skills needed]

### Execution Order
1. [Task 1, Task 3] (parallel -- no dependencies)
2. [Task 2] (depends on Task 1)
3. [Task 4] (depends on Task 2, Task 3)

### External Dependencies
- [dependency] -- [who owns it] -- [current status]

Rules

• Every task must have at least one acceptance criterion that can be empirically verified
• Do not create tasks that cannot be verified -- if you cannot define how to prove it is done, the task is not well-scoped
• Every Task / Bug / Sub-task / Improvement is scoped to exactly one repo -- if the work spans repos, split into per-repo work units under a shared parent Story (see step 1.5)
• Any task proposing a new material dependency names its trust class, states that class's required evidence and whether human ratification is needed, and updates .lisa/DEPENDENCY_DECISIONS.md in the same change (see step 4.5) -- a proposed material dependency with no named class is not ready to build
• Any task removing, replacing, or internalizing a material dependency carries all seven confidence-rebuild kit criteria -- corpus, conformance, negative fixtures, coverage-as-gap-detector, provenance/license, migration/update plan, and rollback criteria (see step 4.6) -- unless it explicitly justifies why the dependency is non-material; a within-trust-class version bump does not carry the kit
• Keep tasks ordered so that no task references work that has not been completed by a prior task
• Flag any task that requires access, permissions, or external input not yet available
• Prefer more small tasks over fewer large tasks -- smaller tasks are easier to verify and less risky to fail
• Do not create placeholder or "TODO" tasks -- every task should describe concrete work