Wr plan
🧠Personal toolkit for repo-local AI agent workflows
npx -y skills add nbsp1221/wayrail --skill wr-planAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use when converting a ready wayrail spec.md into an implementation plan.md. Reads a selected spec item, checks Planning Handoff readiness, performs bounded repository research, and writes only plan.md. Do not use for code implementation, verification evidence, review conclusions, creating tasks.md, or editing spec.md.
SKILL.md
3.2 KB, 613 tokens by cl100k_base, as published. Nobody here has run it
wr-plan
Create the implementation plan for a ready wayrail spec item.
Use When
- The user asks to run
wr-plan, create a plan, or move a ready wayrail spec item from spec stage to plan stage. - The input includes an explicit spec item path or an unambiguous current spec item from the conversation.
- The target spec item has
spec.mdandplan.md.
Do Not Use When
- The user asks to start a new spec item; use
wr-spec. - The user asks to implement code.
- The user asks to verify completed work.
- The user asks to review an implementation.
- The spec is blocked before planning.
- The request would create
tasks.md.
Inputs
Prefer an explicit spec item path such as specs/<id>/spec.md.
If the conversation has exactly one unambiguous current spec item, use that path and state it.
If neither is available, ask for the spec path.
Readiness Gate
Read spec.md before writing anything.
Check Planning Handoff.
Stop without writing when the handoff says Blocked before wr-plan.
Stop without writing when Resolve Before Planning contains unresolved questions.
Workflow
- Identify the target spec item.
- Read
spec.md. - Confirm the readiness gate.
- Read the problem, requirements, success criteria, scope, constraints, assumptions, and open questions.
- Do bounded repository research:
- inspect relevant docs
- inspect likely target files
- identify local implementation and test conventions
- identify likely verification commands
- Classify unknowns before asking the human.
- Write only
plan.md. - Report the plan path and recommend implementation as the next action.
Plan Authoring Rules
Write plan.md with these sections:
- Overview
- Requirements Trace
- Scope
- Context
- Decisions
- Implementation Units
- Verification
- Risks
- Implementation Handoff
Trace every relevant requirement or success criterion to plan coverage, or explicitly mark it out of scope with rationale. Use coarse implementation units, not patch-level tasks. Each implementation unit should include:
- Requirements:
- Files:
- Depends on:
- Approach:
- Verification:
Unknown Handling
Classify unknowns before asking the human.
Human judgment is required for product behavior, scope, quality bar, security, privacy, or risk-tolerance decisions that cannot be resolved from the repository.
Technical convention questions should be resolved from repository context or focused research.
Execution-time uncertainties belong under Risks.
Boundaries
Write only plan.md.
Do not edit source code.
Do not run implementation.
Do not collect verification evidence.
Do not write review conclusions.
Do not create tasks.md.
Phase 1 does not require a script for wr-plan; planning is judgment-heavy.
Completion
Report the written plan.md path.
State whether implementation can begin.
Do not claim verification or review has happened.
What ships with it: 1 file
180 B alongside SKILL.md
agents/
- openai.yaml180 B
Gives 0 of the 12 instructions most plan spec skills give in 613 tokens
Counted across 1,099 of the 1,860 authors here whose files we hold, read 2026-08-07
- Ask one question at a timein 51 of 1099
- Break plans into vertical slicesin 29 of 1099, across 11 files
- Publish issues in dependency orderin 27 of 1099, across 9 files
- Iterate until user approves the breakdownin 25 of 1099, across 7 files
- Explore the repository to understand the codebase statein 24 of 1099, across 7 files
- Use domain glossary vocabularyin 23 of 1099, across 5 files
- Apply correct triage labels to published issuesin 23 of 1099, across 5 files
- Prefer AFK slices over HITLin 22 of 1099, across 7 files
- Write a specification before writing any codein 22 of 1099, across 14 files
- Write failing tests before implementation codein 22 of 1099, across 20 files
- Ask clarifying questions until requirements are concretein 21 of 1099, across 13 files
- Respect existing architecture decision recordsin 20 of 1099, across 5 files
Said here and by no other author read
- stop without writing when readiness is blocked
- stop without writing on unresolved planning questions
- trace every requirement to plan coverage or mark out of scope
- use coarse implementation units
- classify unknowns before asking the human
- resolve technical convention questions using repository context
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.