Help us improve
Share bugs, ideas, or general feedback.
From lockedin
Generate new project ideas, side-project pitches, or career-move proposals from the LockedIn vault. The skill walks the user's past work, identifies under-exploited combinations of skills and domains, and proposes concrete next moves with a one-paragraph rationale grounded in the user's actual experience. Two-turn writer/reviewer pattern; reviewer scores feasibility, novelty, evidence ground, scope match, and motivation alignment. Use when the user says any of <!-- ko-example -->"what should I work on next", "ideas for side projects", "새 프로젝트 아이디어", "next move", "career pivot ideas", "흥미로운 조합"<!-- /ko-example -->, or asks for proposals derived from their own experience. Output: 3 to 5 ideas in markdown, each one paragraph long with cited entities resolved to natural language, plus a JSON rubric score for the set.
npx claudepluginhub daypunk/lockedin --plugin lockedinHow this skill is triggered — by the user, by Claude, or both
Slash command
/lockedin:lockedin-render-ideasThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Status: **v1.1 (skeleton)**. Skill scaffold ships with prompts and a
Brainstorms and explores ideas for projects or features using subagents. Useful for generating creative solutions or exploring possibilities.
Identifies unique specific knowledge from curiosity and passion to build irreplaceable careers. Use when choosing skills to develop, differentiating, or fearing replacement.
YC Office Hours brainstorming: startup mode with six forcing questions to validate demand, or builder mode for design thinking on side projects. Saves a design doc. Use before planning reviews.
Share bugs, ideas, or general feedback.
Status: v1.1 (skeleton). Skill scaffold ships with prompts and a minimal rubric. Cross-source calibration is a v1.2 target. The skill is functional today; the rubric has not been validated against an external corpus the way jaso and resume-en have.
render-resume-en /
render-jaso.render-interview./lockedin init or ingestion.Writer turn surfaces 3 to 5 ideas. Reviewer turn re-loads RUBRIC.md
fresh and scores the set on five dimensions.
3 to 5 ideas, each one paragraph long, with a one-sentence pitch
followed by two to three sentences of rationale that cite vault
entities. Slug citations ([[type/slug]]) are resolved to natural
language by the orchestrator before the user sees the artifact.
SKILL.md
prompt-writer.md writer-turn instruction
prompt-reviewer.md reviewer-turn instruction (separate Claude turn)
RUBRIC.md 5-dimension scoring contract (calibration v1.2)
The rubric here is v1.1 foundational. Cross-source calibration (against entrepreneur / career-coach corpora) is the v1.2 target. Until then, the reviewer turn is self-consistent against this rubric, but does not match the calibration density of the jaso or resume-en rubrics.