cracked v1.3.0

Everything from the improvement review, in four releases (v1.2.0 → v1.3.0):

• Honest grading: /mock-interview stress-tests your code against a brute force on thousands of random inputs; /oa-sim keeps hidden tests outside your folder and uses per-language time limits.
• Works everywhere: a portable code runner (macOS, Linux, Windows) and a fallback for agents without a Skill tool.
• Fresh facts: a tested CTC calculator for FY 2026-27 (both regimes, labour-code PF), 200-character LinkedIn notes, no hard-coded years.
• Better company data: aliases like "Tata Consultancy Services" → TCS, automatic window widening, and a freshness date on every result.
• Full eval coverage: 21 cases, one or more per skill, all passing.

Full details in CHANGELOG.md.

cracked v1.1.0

Neofetch-style profile READMEs with your photo as ASCII art, a zero-dependency image-to-ASCII converter, and animated demos for six skills. See CHANGELOG.md.

cracked v1.0.0

cracked v1.0.0: your AI interviewer, inside your terminal

18 agent skills that turn Claude Code, Codex, Cursor, Copilot or Gemini CLI into a full placement-prep team.

Highlights

• 🏢 Company mode, 470+ companies. Say "mock interview for Google" or "Flipkart OA" and every skill adapts, using frequency-ranked, community-reported LeetCode questions (30-day / 3-month / 6-month / all-time), reported rounds, published values and resume bars.
• 🎤 Mock interviews that feel real. Withheld constraints, a 4-rung hint ladder, your code run against hidden edge cases, per-phase monitoring of how you react, voice mode (/voice in, OS text-to-speech out), and integrity checks.
• ⏱️ OA simulator with hidden tests. Real files you solve in your editor, graded AC / WA / TLE / RE against tests verified by an executed reference solution.
• 📄 Brutal, calibrated resume roast. An ATS score out of 100 with its breakdown, judged against your level and market, and rewrites that never invent a single metric.
• 🧭 A journal that remembers. Every session logs a shareable scorecard to ~/.cracked/; /prep-plan and /rejection-review turn your real weak spots into your next steps.

All skills

/setup-cracked · /company-intel · /resume-blueprint · /mock-interview · /oa-sim · /pattern-coach · /core-viva · /system-design-mock · /hr-round · /defend-my-project · /roast-my-resume · /jd-match · /github-glowup · /make-it-hireable · /referral-dm · /prep-plan · /rejection-review · /offer-decode

Install

# Claude Code
claude plugin marketplace add nishant-Tiwari24/cracked
claude plugin install cracked@cracked

# Codex
codex plugin marketplace add nishant-Tiwari24/cracked
codex plugin add cracked@cracked

# Cursor, Copilot, Gemini CLI, OpenCode, Windsurf, Amp…
npx skills@latest add nishant-Tiwari24/cracked

Quality

• claude plugin validate --strict passes; the structural validator runs in CI on every PR.
• A behavioural eval suite of 12 cases, graded by deterministic checks plus LLM judges: 12/12 pass, +0.38 over the same model without the plugin.

Company question data comes from liquidslr/leetcode-company-wise-problems, fetched at runtime with thanks.