templates/skills/dev/analysis/SKILL.md
Create structured analyses with numbered findings, execution plans, and task materialization
npx skillsauth add hivellm/rulebook analysisInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Create a structured analysis for: $ARGUMENTS
Analyses live in docs/analysis/<slug>/ and are ALWAYS split into
numbered files, one theme per file, in reading order:
docs/analysis/<slug>/
├── README.md # index + executive summary; links every numbered file
├── 01-<theme>.md # one theme per file (e.g. 01-measurements.md)
├── 02-<theme>.md # (e.g. 02-root-causes.md)
├── ...
└── NN-execution-plan.md # last file, only when the analysis proposes work
Never put the whole analysis in a single file. Findings are numbered F-001..F-NNN globally across the analysis (numbering continues from one file to the next), each with: title, evidence (file:line), impact, confidence.
Steps:
docs/analysis/<slug>/ following the structure aboverulebook_memory {kind:"knowledge"|"learning", action:"list"}NN-execution-plan.mdREADME.mdrulebook_memory {action:"add"} tagged analysis:<slug>rulebook_task {action:"create"}research
Author a rulebook task spec interactively — research, draft, ask the user clarifying questions, confirm, then create the tasks in rulebook ready for /rulebook-driver. Use when the user wants to plan/spec a feature before implementing.
development
Behavioral guidelines to reduce common LLM coding mistakes — overcomplication, sloppy refactors, hidden assumptions, weak goals. Use when writing, reviewing, or refactoring code. Auto-applies; invoke explicitly via /karpathy-guidelines or 'follow karpathy discipline'.
development
Spec-driven task management for features and breaking changes with OpenSpec-compatible format
tools
Tool: Anthropic Claude Code CLI (`npm install -g @anthropic-ai/claude-code`)