How this skill is triggered — by the user, by Claude, or both
Slash command
/rp-mini:rp-buildThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this when the user wants code changes and the task benefits from curated context.
Use this when the user wants code changes and the task benefits from curated context.
Check workspace binding first. On mismatch, pass root=<absolute path> on every rp-mini tool call.
If a shell is available and no MCP client is loaded, use node packages/server/dist/cli.js tool <workspace> <tool> --json-args '...' or its wrappers.
get_file_tree, file_search, or get_code_structure.context-builder subagent with:
planrewrite unless the user requested augment or preserve<questions> plus a saved profile name, read the cited files, check git history or searches, and answer only conclusive questions with <answer key="..." source="orchestrator" >...</answer>. Escalate with AskUserQuestion only for blocking AND high-stakes questions: irreversible/destructive actions, product policy, money/auth/data-loss. Append <answers> with prompt op=append, resume with manage_selection op=load_profile, and remember advisory questions never interrupt.Do not skip context-builder for non-trivial code work; the quick scan is orientation, not the deep exploration.
npx claudepluginhub apollostreetcompany/rp-mini --plugin rp-miniCreates bite-sized, testable implementation plans from specs or requirements, with file structure and task decomposition. Activates before coding multi-step tasks.