| name | plan-proposal |
| description | Write a detailed implementation plan for an approved Tensor4all.jl spec |
Plan Proposal
You are helping a contributor write an implementation plan for a Tensor4all.jl
issue whose spec has been approved (spec_approved label).
Before you start
- Read
CONTRIBUTING.md for the contribution flow.
- Read
AGENTS.md for codebase conventions and architecture.
- Read the approved spec on the issue.
- Explore the codebase in depth: relevant source files, existing tests,
module boundaries.
Your task
Draft a detailed implementation plan to post as an issue comment.
Output format
Write the plan with these sections:
## Implementation Plan
### Overview
(1-2 sentences: what this plan delivers)
### Task 1: [name]
**Affected files:** (list modules/files, no line numbers needed)
**What to do:** (describe the change)
**Tests:** (what to test)
### Task 2: [name]
...
### Task ordering
(Which tasks depend on others, suggested sequence)
### Test strategy
(Overall approach: unit tests, integration tests, edge cases to cover)
### Open questions
(Anything that needs maintainer input before implementation)
Guidelines
- Break work into small, independently testable tasks.
- Each task should touch a coherent set of files in one module.
- Describe what to change and why, not exact code. The implementer will write
the code.
- Note any tasks that require C API changes (these need a separate
tensor4all-rs PR first — see cross-repo dependency rules in
AGENTS.md).
- Reference the acceptance criteria from the approved spec. Every criterion
should map to at least one task.
- Prefer TDD: mention which tests to write before which implementation.