Stage 02
Build your proof
Portfolio projects, GitHub profile, MCP servers, and the contribute-to-hire playbook.
Portfolio projects: building one strong proof-point repo
Why one well-built, deployed, blog-backed portfolio project beats five half-done ones — how to pick it, scope it, and ship it as the single repo every recruiter clicks first.
Your GitHub profile README: what recruiters actually see
The GitHub profile README structure that converts recruiter visits to outreach — what to put above the fold, what pinned repos to show, and what to delete (stats cards, snake graphs).
Repo READMEs that make a recruiter open the link
The repo README structure that converts a one-second skim to a click — opening hook, screenshot, the 6 sections, and why your stack badges are killing the click-through rate.
Building in public for inbound recruiter DMs
The build-in-public cadence (Twitter/X + LinkedIn + blog) that turns shipped work into inbound recruiter DMs — the post formats that convert and the ones that don't.
Contribute-to-hire: finding repos that hire from India
How to find OSS repos whose maintainers convert contributors into hires — the GitHub signals (Series A-C, India-friendly team page, contributor-to-employee history) and the search queries that surface them.
Contribute-to-hire: picking a winnable issue (verify rules + -linked:pr)
How to pick an open-source issue you can actually merge — the verify-first checklist (no assignee, no open PR, no backend blocker) and the GitHub `-linked:pr` search trick.
Contribute-to-hire: writing a PR that gets merged
How to write an open-source PR that gets merged the first time — small surface area, tests included, the PR description structure, and how to handle maintainer review without ego.
Contribute-to-hire: turning a merged PR into an offer
The post-merge play that converts an OSS contribution into a job offer — the timed DM to the maintainer, the portfolio framing, and the interview-skipping referral path.
Reading a giant OSS codebase fast
How to orient yourself in a 200k-line OSS codebase in 2 hours — the entrypoint walk, the test-first map, AST grep, and the LLM-assisted code tour that actually works.
Getting a complex OSS repo running locally
How to get a multi-service OSS repo (monorepo, Postgres, Redis, Docker) running locally on day 1 — the dev-container shortcut, the README-vs-reality gap, and the unstuck checklist.
Working with OSS maintainers without getting ghosted
How to communicate with open-source maintainers so they actually respond — the right channel, the right cadence, what to never ask, and the trust-building micro-moves that lead to hire conversations.
Building an MCP server as a portfolio weapon
How to build a real Anthropic Model Context Protocol server end-to-end as a portfolio project — tool schemas, stdio vs SSE transport, deployable demo, and the blog post that gets inbound.