ai-native-camp/camp-1
7 skills9.4k installs1.7k starsGitHub
Install
npx skills add https://github.com/ai-native-camp/camp-1Skills in this repo
1Day1 Onboardingday1-onboarding is a journey-wide camp skill that walks a solo builder through standing up Claude Code on day one of an AI-native curriculum. It prioritizes terminal-first workflow: curl or irm installers, no Node.js requirement, first `claude` invocation, and subscription login checks. Optional editor setup (Antigravity, Cursor, VS Code) is framed as helpful but not mandatory, which matches how many indies start with a single terminal tab. The block also teaches practical environment literacy—revealing hidden folders on macOS, opening the project in Finder, and noticing `.claude/skills/` where camp materials live. Triggers align with onboarding, install help, and first-run confusion across Idea discovery and Build agent-tooling, because you can return to it whenever a new machine or teammate needs the same baseline. Complexity is beginner; deliverable is a working agent CLI session ready for later camp blocks.1.6kinstalls2Day2 Create Context Sync SkillDay 2 Block 0 from AI Native Camp walks solo builders and small teams through choosing everyday tools and packaging a Context Sync agent skill. Context Sync means Claude Code gathers overnight Slack threads, unread mail, calendar blocks, Notion or Linear tasks, and similar sources into one structured morning document instead of manually hopping between five apps. The material explains what context switching costs in time, how skills generalize repeated work into a one-line invoke, and how to pick two or three integrations based on which app you open first at work. It is curriculum-style guidance rather than a finished marketplace skill, but the pattern maps cleanly to indie operators who live in chat, mail, and issue trackers. Use it when you want a concrete template for multi-MCP or multi-connector morning rituals and when you are ready to author SKILL.md triggers like “sync my morning context.”1.6kinstalls3Day2 Supplement McpDay2 Supplement MCP is a Korean-language concept primer for solo builders and camp learners who need a shared mental model before touching MCP configuration in Claude Code. It defines MCP as the open Model Context Protocol that links AI apps to external tools, using USB-C and translator analogies so heterogeneous services speak one contract. The material contrasts ad-hoc integrations with a single MCP server layer serving Slack, Notion, and Calendar uniformly, then names Host, Client, and Server roles with a simple architecture diagram. It is not a deploy script; it is structured EXPLAIN content that camp facilitators print at Phase A start with mandated official documentation URLs. Use it when you are new to agent tooling and want textbook clarity on why MCP exists and how pieces fit, immediately before hands-on server setup. Skip if you already run production MCP servers and only need troubleshooting or security hardening guides.1.5kinstalls4Day4 Wrap And Analyzeday4-wrap-and-analyze is a camp-style agent skill for solo builders using Claude Code who want structured end-of-session hygiene without manually summarizing hours of work. It explains multi-agent orchestration through a concrete session-wrap metaphor: four parallel experts each analyze docs, repetitive automation opportunities, lessons learned, and next-week tasks, then a second phase runs duplicate-checker to merge overlapping recommendations. The pattern reinforces sub-agent Task usage from earlier lessons but shifts from sequential to parallel execution for speed. Ideal when you finish a heavy coding day and need consistent artifacts—updated documentation pointers, automation ideas, learning notes, and a deduplicated action list—without writing the synthesis yourself. Complexity is intermediate because it assumes comfort spawning sub-agents and interpreting combined outputs. It is educational content (including Korean explainers) aligned with ai-native-camp curriculum rather than a production integration.1.4kinstalls5Day5 Fetch And Digestday5-fetch-and-digest is a camp-style agent skill that explains how to build a content ingestion factory for Claude Code: fetch external URLs, translate in layers, then digest with quizzes. It maps fetch-tweet and fetch-youtube to delivery steps and content-digest to retention-style learning, mirroring a production line from raw URL to owned knowledge. Solo builders use it while learning skill authoring—when they want Twitter, YouTube, or article workflows without rebuilding prompts each time. The documented order avoids jumping to full translation before summary and insights, which keeps token use and attention on what matters. After mastery, the same pattern extends to newsletters, docs, or competitor pages in Grow and Validate research.1.3kinstalls6Day6 Prd Submitday6-prd-submit is a structured agent skill for AI Native Camp participants—and any solo builder mirroring that curriculum—who must turn an idea into a reviewable Product Requirements Document and submit it through GitHub. The skill walks through environment verification (git and gh), confirms GitHub identity, then drives PRD authoring against an explicit markdown template emphasizing one-line problem statements, quantitative current state, a one-week desired picture, and numeric success criteria. After format validation, it guides branch, commit, push, and Pull Request creation so instructors or teammates can approve scope before Build.work begins. The copy is Korean-forward with plain-language definitions for repo, branch, and PR, which lowers friction for first-time git users. It is a workflow skill with hard gates on tooling rather than a generic writing assistant, and it assumes collaborator access on a camp repository.1.2kinstalls7My Fetch TweetMy Fetch Tweet is a Korean-first agent skill that turns an X/Twitter link into structured intelligence for solo builders monitoring the timeline. It normalizes the URL to FxEmbed’s API, pulls tweet text (with expanded URLs), author metadata, engagement counts, media, and quoted posts, then deliberately stages output: a fast Korean summary with author and metrics, three insight lenses (core message, industry implication, personal takeaway), and only then a full natural translation with quoted tweets and glossed terms. Triggers match everyday Korean requests like “트윗 번역” or “트윗 가져와.” It excels in Idea research when you are tracking builders and investors, and in Grow content when you repurpose discourse for newsletters or threads. It does not schedule posts, run bulk analytics, or replace official X API compliance reviews.744installs