newgrad-ops

Career-Ops for students trying to get their first tech offer.

Turn Claude Code, Codex, Gemini CLI, Cursor, or any AI coding assistant into an internship and new-grad job-search command center.

Track applications · tailor resumes · prep for online assessments · grind company-specific LeetCode · audit your GitHub portfolio · manage referrals · build a weekly recruiting plan.

Not another job tracker — an agentic career operating system for breaking into tech.

---

One student journey

/student company anthropic
/student tailor-resume https://anthropic.com/careers/software-engineer
/student interview weak-patterns
/student portfolio-audit
/student alumni-message anthropic

Output:

────────────────────────────────────────────────────────────
Anthropic New-Grad Prep Pack

APPLICATION
  Resume focus: Python + reliability signals, AI tooling projects
  Fit tip: reframe projects toward evals and observability

INTERVIEW FORMAT (from anthropic.com/careers)
  Tools: Colab · CodeSignal
  Lookups: allowed, but be fluent in syntax
  AI usage: only when explicitly permitted

OA PREP
  Focus: Sliding Window · Graphs · DP
  Problems: Longest Substring · Course Schedule · Word Break

BEHAVIORAL
  Themes: ownership · judgment · mission alignment
  Stories: shipped without requirements · debugged the unknown · disagreed productively

WEEKLY PLAN
  Mon: 5 sliding window problems
  Tue: resume tailoring pass (Anthropic)
  Wed: 5 graph problems
  Thu: mock CodeSignal session (45 min, plain Python)
  Fri: alumni outreach × 2

────────────────────────────────────────────────────────────

---

Modules

Module	Commands	What it does
Apply Ops	/student apply-plan, fit-score, tailor-resume, track, follow-up	Application planning, fit scoring, resume diffs, tracking
Interview Ops	/student interview diagnose, weak-patterns, review-today, explain-with-links	Pattern diagnosis, spaced review, company problems
Company Ops	/student company <name>	Full prep pack: format, OA patterns, behavioral, weekly plan
Portfolio Ops	/student portfolio-audit, github-readme, project-match, project-upgrade-plan	GitHub audit, README upgrades, project-role matching
Referral Ops	/student referral-map, alumni-message, recruiter-dm, follow-up-referral	Alumni connections, ethical outreach drafts

---

Problem lists

Three curated lists for structured interview prep:

List	Problems	Source
NeetCode 150	150 problems, 16 patterns	neetcode.io — free video explanations
Blind 75	76 problems (highest-frequency)	neetcode.io
Monster 50	50 highest-ROI problems, 12 categories	algo.monster/practice — organized by cue and trap

All three are mapped in data/interview/ with patterns, traps, and public solution links.

---

Quick start

1. Clone the repo

git clone https://github.com/[you]/newgrad-ops.git

2. Point your AI CLI at the skills directory

For Claude Code:

# In your CLAUDE.md or via /config, add the skills path:
# skills_dir: ./skills
claude
/student company anthropic

For Codex CLI:

codex --skills-dir ./skills

For Cursor: add skills/student-career/SKILL.md to your context.

3. Use the scripts locally

python scripts/score_resume.py my-resume.md
python scripts/score_readme.py ~/projects/my-project/README.md
python scripts/generate_review_schedule.py solved.json
python scripts/search_problems.py --pattern "Graphs" --difficulty Medium

---

Company prep packs

Built-in coverage for:

• Anthropic — live escalating challenge (Colab/CodeSignal), AI usage policy
• OpenAI — practical/debugging style, not blank-file LeetCode
• Google — Google Docs coding, no IDE, classic LeetCode-style
• Meta — formerly Facebook — speed-focused, classic LeetCode-style
• Amazon — classic LeetCode-style + Leadership Principles as their own bar
• Microsoft — classic LeetCode-style with a behavioral warm-up each round
• Apple — DP/sliding-window take-home filter, memory/defensive-coding emphasis
• Netflix — behavioral weighted heaviest of any company here, coding weighted least
• Stripe — take-home projects, code quality focus, not LeetCode-style

Each file states its Coding style explicitly — some of these companies don't run a LeetCode-style interview at all, and treating every "coding interview" the same way is the fastest way to over- or under-prep.

Add your own: create data/companies/<name>.md following the same template (see CONTRIBUTING.md).

---

Safety principles

• No auto-apply — all applications are reviewed and submitted by you
• No auto-email — all outreach messages are drafts; you send them manually
• No scraped paid content — links go to public pages only (NeetCode free, AlgoMonster free tier, LeetCode problem statements)
• No fake referrals — referral templates are ethical, one-at-a-time, manually sent
• No destructive file writes — scripts print to stdout; nothing is written without your intent
• Company prep packs cite public sources — interview format data comes from public careers pages

---

Examples

See the examples/ directory for realistic sample outputs:

• anthropic-prep-pack.md — full company prep pack
• new-grad-application-plan.md — weekly application plan
• weak-pattern-report.md — pattern coverage and drill list
• portfolio-audit.md — before/after GitHub audit

---

Contributing

See CONTRIBUTING.md for the full guide, ground rules, and a list of good first issues (the Blind 75 and NeetCode 150 maps both have known gaps — easy, well-scoped PRs).

Quick version: to add a company pack, create data/companies/<name>.md using the existing files as a template and cite only public sources. To add problems to the maps, edit the JSON files in data/interview/ — include pattern, difficulty, a verified public URL, and at least one trap.

By participating you agree to the Code of Conduct.

---

License

MIT