From rnd-framework
Use when creating git commits — enforces commit message style, length limits, and user confirmation before committing
How this skill is triggered — by the user, by Claude, or both
Slash command
/rnd-framework:committingThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
1. **One line, imperative mood.** "Fix race condition in token refresh", not "Fixed" or "Fixes" or "Fixing".
i < len to i <= len".git remote get-url origin before using platform-specific CLI tools (e.g., gh is GitHub-only). For PRs/MRs, ask the user or infer from the remote URL.Use only git commands and dedicated Claude Code tools — use Read instead of cat, Grep instead of grep, Edit instead of sed, Glob instead of find. Prefer dedicated tools over shell equivalents — they are reviewable and produce cleaner diffs.
git status and git diff (separate Bash calls, never chained with &&).cat.grep.git log --oneline -5 to match the repo's commit style.AskUserQuestion before committing.git add <specific files> — never git add -A or git add ..git commit -m "..." in a separate Bash call.Always use AskUserQuestion to confirm the commit message before running git commit. Present 2-3 message options with the best one marked "(Recommended)".
Example:
AskUserQuestion:
question: "Commit message?"
options:
- label: "Add Write tool to verifier agent (Recommended)"
description: "Describes the functional change"
- label: "Fix verifier using bash heredocs for file creation"
description: "Describes the problem that was solved"
Add rate limiting to public API endpoints
Fix duplicate webhook delivery on retry
Remove deprecated v1 auth endpoints
Extract PDF parser into standalone module
Update Stripe SDK to v15 for SCA compliance
Update files # vague
Fix bug # what bug?
Refactor code for better quality # says nothing
Add AI-powered verification system # leaks internals
Various improvements and fixes # filler
Change line 42 in auth.ts # describes how, not what
npx claudepluginhub oleksify/rnd-framework --plugin rnd-frameworkProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.