From flowstate
Generate a detailed implementation plan for a feature or change before coding begins. Use when the user says "plan this", "how should we implement", "design approach", or when a task needs architectural thinking.
How this skill is triggered — by the user, by Claude, or both
Slash command
/flowstate:ideasonnetThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Create a detailed implementation plan for a feature or change, saved for later review.
Create a detailed implementation plan for a feature or change, saved for later review.
Feature description (optional): $ARGUMENTS
Verify .backlog/ exists. If not, tell the user to run /flowstate:setup first.
If $ARGUMENTS provided, use it as the starting point. Otherwise ask:
Before exploring code, gather backlog context:
node "${CLAUDE_PLUGIN_ROOT}/dist/bin/flowstate.js" learning-search --query "{{FEATURE_DESCRIPTION}}" --limit 3 --json true
The CLI returns only active learnings, scored by keyword relevance. Use title and reasons to assess relevance. Read the full learning file only for high-scoring matches that directly affect the plan..backlog/tasks/active/ — the plan should account for work already in progress to avoid conflicts or duplication..backlog/reports/pending/ for related bugs or findings that the plan should address or acknowledge.If no matches found, skip silently.
Based on the description:
cat <<'BODY' | node "${CLAUDE_PLUGIN_ROOT}/dist/bin/flowstate.js" idea-create --title "{{TITLE}}" --complexity {{COMPLEXITY}} --body -
{{PLAN_CONTENT}}
BODY
The CLI handles ID assignment, file creation, and placement in ideas/pending/.
Complexity guidelines:
Created PLN-{{ID}}: {{TITLE}}
Complexity: {{COMPLEXITY}}
File: .backlog/ideas/pending/PLN-{{ID}}-{{slug}}.md
Next: /flowstate:review-idea PLN-{{ID}} — Approve, discard, or revise
npx claudepluginhub jmlweb/skills-monorepo --plugin flowstateCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.