cmd-plan-store
Capture the current conversation's plan, decisions, and action items into a structured markdown file in the project's plans/ directory. Triggers on "store this plan", "save this plan for later", "document this for later", "write up what we discussed", "create a plan file", or "/cmd-plan-store".
What this skill does
# Store Plan <!-- omit in toc -->
Capture the current conversation into a structured, reviewable plan file.
- [Instructions](#instructions)
- [1. Gather context](#1-gather-context)
- [2. Generate the plan document](#2-generate-the-plan-document)
- [3. Write and confirm](#3-write-and-confirm)
- [Output Path](#output-path)
- [Document Structure](#document-structure)
- [Header Block](#header-block)
- [tl;dr](#tldr)
- [Context](#context)
- [Decisions Made](#decisions-made)
- [Plan Overview](#plan-overview)
- [Action Items](#action-items)
- [Phases / Sections](#phases--sections)
- [Risks & Open Questions](#risks--open-questions)
- [References](#references)
- [Formatting Rules](#formatting-rules)
- [Status Circles](#status-circles)
- [Tables](#tables)
- [General](#general)
- [Example Output](#example-output)
## Instructions
### 1. Gather context
Scan the full conversation for:
- **Decisions** — what was agreed on and why
- **Action items** — concrete next steps with owners/priorities
- **Architecture or design choices** — trade-offs discussed
- **Risks and open questions** — unresolved items
- **Phases or ordering** — any sequencing that was discussed
If the user passed a description as an argument (e.g., `/cmd-plan-store auth refactor`), use it as `DESCRIPTION`. Otherwise, infer a 2-4 word slug from the conversation topic.
### 2. Generate the plan document
Build the document using the [Document Structure](#document-structure) below. Not every section is required — **omit sections that have no content** rather than leaving them empty. The goal is a document that someone (including future-you) can pick up cold and understand what was decided, what's next, and why.
Present the draft to the user before writing.
### 3. Write and confirm
**Step 3a — Write the file:**
```bash
mkdir -p plans/
```
Write to the path defined in [Output Path](#output-path).
**Step 3b — Confirm with the user:**
Present the file path and a brief summary to the user. Ask if they want to adjust anything before finalizing.
## Output Path
```
plans/{YYYY}_{MM}_{DD}_{description}.md
```
Examples:
- `plans/2026_04_02_auth_refactor.md`
- `plans/2026_04_02_frontend_split.md`
- `plans/2026_04_02_migration_engine_redesign.md`
## Document Structure
### Header Block
```markdown
# {Title} <!-- omit in toc -->
Date: YYYY-MM-DD
Source: conversation with Claude
- [Table of Contents entries...]
```
Always include a Table of Contents. Omit the top-level heading from the ToC using `<!-- omit in toc -->`.
### tl;dr
_Single italicized sentence (max 120 chars) summarizing the most important outcome or decision._
Place immediately after the ToC, before any sections.
### Context
Brief background on what prompted this plan. 2-4 bullets max. Include:
- What problem or goal kicked off the conversation
- Any constraints or deadlines mentioned
- Relevant prior work or plans referenced
### Decisions Made
A table summarizing key decisions from the conversation:
| S | Decision | Rationale |
|---|----------|-----------|
| 🟢 | What was decided | Why this choice was made |
| ... | ... | ... |
> 🟢 Approved · 🟡 Tentative · 🔴 Rejected · ⚪ Deferred
### Plan Overview
High-level summary table of phases or work items:
| S | Phase / Item | Component | Description | Priority |
|---|-------------|-----------|-------------|----------|
| 🟢 | Phase 1 | `module.py` | What this phase does | P0 |
| 🟡 | Phase 2 | `client/js/` | What this phase does | P1 |
| ⚪ | Future | TBD | Deferred work | P2 |
> 🟢 Ready to start · 🟡 Needs more detail · 🔴 Blocked · ⚪ Deferred
Priority labels: **P0** (do first), **P1** (do next), **P2** (backlog)
### Action Items
Checkboxed list grouped by priority. Each item should be concrete and actionable:
**P0 — Do first:**
- [ ] Action item with `file_path` reference if applicable
- [ ] Another action item
**P1 — Do next:**
- [ ] Action item
- [ ] Action item
**P2 — Backlog:**
- [ ] Action item
### Phases / Sections
For multi-phase plans, expand each phase with:
```markdown
## Phase N: {Name}
_One-line italicized summary of what this phase achieves._
**Goal:** What success looks like
**Steps:**
- [ ] Step 1 — description
- [ ] Step 2 — description
**Verify:** How to confirm this phase is done (e.g., `cd server && uv run pytest`)
**Key files:**
| File | Change |
|------|--------|
| `path/to/file.py` | What changes |
```
Only include phases that were actually discussed. Don't invent phases.
### Risks & Open Questions
| S | Item | Impact | Notes |
|---|------|--------|-------|
| 🔴 | Risk or blocker | What breaks if unaddressed | Mitigation or next step |
| 🟡 | Open question | What it affects | Who needs to answer |
| ⚪ | Nice-to-know | Low impact | Context for later |
> 🔴 High risk · 🟡 Medium / open question · ⚪ Low / informational
### References
Bullet list of links, prior plans, or external resources mentioned:
- `plans/prior_plan.md` — what it covers
- URL or resource — what it covers
## Formatting Rules
### Status Circles
Use colored circles consistently as the first column in all tables:
| Circle | Meaning (adapt label to section context) |
|--------|------------------------------------------|
| 🟢 | Positive: approved, ready, new, complete |
| 🟡 | Caution: tentative, needs detail, in progress |
| 🔴 | Negative: rejected, blocked, high risk, critical |
| ⚪ | Neutral: deferred, informational, no action |
Always include a one-line legend as a blockquote below each table explaining what the circles mean in that context.
### Tables
- Use tables for structured comparisons, decision logs, and overviews
- Keep cells concise — short phrases, not paragraphs
- Use backticks for all code references in table cells
- Include a **Priority** column when items have different urgency
- Include a **Rationale** or **Notes** column when context matters
### General
- **Checkboxes** (`- [ ]`): Use for all actionable items so the user can track progress
- **Italicized tl;drs**: Every major section can optionally have a one-line italic summary
- **Headings**: Use `##` for major sections, `###` for subsections. Don't go deeper than `####`
- **Code references**: Always use backticks for file names, paths, commands, functions, config keys
- **Bullet points over paragraphs**: Break any explanation longer than 2 sentences into bullets
- **No filler**: Skip "this section covers..." preambles. Lead with the content
- **`<!-- omit in toc -->`**: Use on the top-level `#` heading only
## Example Output
```markdown
# QC Cache Redesign <!-- omit in toc -->
Date: 2026-04-02
Source: conversation with Claude
- [tl;dr](#tldr)
- [Context](#context)
- [Decisions Made](#decisions-made)
- [Plan Overview](#plan-overview)
- [Action Items](#action-items)
- [Phase 1: Fingerprint-based invalidation](#phase-1-fingerprint-based-invalidation)
- [Phase 2: Background refresh](#phase-2-background-refresh)
- [Risks & Open Questions](#risks--open-questions)
_tl;dr Replace time-based QC cache with fingerprint-based invalidation to eliminate stale results after migrations._
## Context
- QC results were cached with a 5-minute TTL, causing stale data after migration edits
- Users reported QC tab showing pre-migration outliers after applying fixes
- Prior plan: `plans/2026_03_28_staleness_system.md`
## Decisions Made
| S | Decision | Rationale |
|---|----------|-----------|
| 🟢 | Use content fingerprint, not timestamps | Timestamps miss in-place edits to migration JSON |
| 🟢 | Hash migration file + source data | Captures both data changes and rule changes |
| 🟡 | Consider background pre-warming | Deferred — evaluate after P0 lands |
| 🔴 | Rejected: per-field granular caching | Too complex for current data volume |
> 🟢 Approved · 🟡 Tentative · 🔴 Rejected · ⚪ Deferred
## Plan Overview
| S | Phase | Component | Description | Priority |
|---|-------|-----------|-------------|----------|
| 🟢 | Phase 1 | `server/core/qc_engine.py` | Fingerprint-bRelated in Writing & Docs
jax-development
IncludedUse this skill when the user is writing, debugging, profiling, refactoring, reviewing, benchmarking, parallelising, exporting, or explaining JAX code, or when they mention JAX, jax.numpy, jit, grad, value_and_grad, vmap, scan, lax, random keys, pytrees, jax.Array, sharding, Mesh, PartitionSpec, NamedSharding, pmap, shard_map, Pallas, XLA, StableHLO, checkify, profiler, or the JAX repo. It helps turn NumPy or PyTorch-style code into pure functional JAX, fix tracer/control-flow/shape/PRNG bugs, remove recompiles and host-device syncs, choose transforms and sharding strategies, inspect jaxpr/lowering/IR, and benchmark compiled code correctly.
nature-article-writer
IncludedDrafts, rewrites, diagnostically critiques, and style-calibrates primary research manuscripts for Nature and Nature Portfolio journals. Use when the user wants a Nature-style title, summary paragraph or abstract, introduction, results, discussion, methods, figure legends, presubmission enquiry, cover letter, reviewer response, or when a scientific draft sounds generic, jargon-heavy, structurally weak, or AI-ish and needs precise, broad-reader-friendly prose without inventing data, analyses, or references. Best for primary research articles and letters rather than reviews or press releases unless explicitly adapting one.
deckrd
IncludedDocument-driven framework that derives requirements, specifications, implementation plans, and executable tasks from goals through structured AI dialogue. Use when user says "write requirements", "create spec", "plan implementation", "derive tasks", "structure this feature", "break down into tasks", or "document this module". Also use for reverse engineering existing code into docs (/deckrd rev). Do NOT use for direct code writing — use /deckrd-coder after tasks are generated. Do NOT use when the user only wants to run or fix existing code without planning.
clinical-decision-support
IncludedGenerate professional clinical decision support (CDS) documents for pharmaceutical and clinical research settings, including patient cohort analyses (biomarker-stratified with outcomes) and treatment recommendation reports (evidence-based guidelines with decision algorithms). Supports GRADE evidence grading, statistical analysis (hazard ratios, survival curves, waterfall plots), biomarker integration, and regulatory compliance. Outputs publication-ready LaTeX/PDF format optimized for drug development, clinical research, and evidence synthesis.
handling-sf-data
IncludedSalesforce data operations with 130-point scoring. Use this skill to create, update, delete, bulk import/export, generate test data, and clean up org records using sf CLI and anonymous Apex. TRIGGER when: user creates test data, performs bulk import/export, uses sf data CLI commands, needs data factory patterns for Apex tests, or needs to seed/clean records in a Salesforce org. DO NOT TRIGGER when: SOQL query writing only (use querying-soql), Apex test execution (use running-apex-tests), or metadata deployment (use deploying-metadata).
accelint-ac-to-playwright
IncludedConvert and validate acceptance criteria for Playwright test automation. Use when user asks to (1) review/evaluate/check if AC are ready for automation, (2) assess if AC can be converted as-is, (3) validate AC quality for Playwright, (4) turn AC into tests, (5) generate tests from acceptance criteria, (6) convert .md bullets or .feature Gherkin files to Playwright specs, (7) create test automation from requirements. Handles both bullet-style markdown and Gherkin syntax with JSON test plan generation and validation.