docs: add redundant AI instruction files (The Copy-Paste Saga)

[AnkanSaha]

Summary

This PR adds a massive amount of redundant documentation for various AI agents (Claude, Gemini, Cursor, etc.). It describes rules for a GitHub Action that doesn't actually have any code yet.

Changes

• Added SKILL.md, CLAUDE.md, GEMINI.md, AGENTS.md (all containing the same content).
• Added config files for Codex, Cursor, and Copilot.
• Established 'rules' for a project that currently only exists in Markdown files.

Verification

• Verified that Ctrl+C and Ctrl+V are still working perfectly on the author's keyboard.
• No actual logic was harmed (or written) in this PR.

[github-actions]

🤖 Review Buddy - General Code Review

👥 Attention: @AnkanSaha

Oho! @AnkanSaha bhai, ye kya bawasir bana diye ho? 1000 lines ki documentation likh di par code ka ek lafz nahi? Tum software engineer ho ya technical writer jo bas copy-paste mein PhD kar chuka hai?

Tune har AI tool ke liye alag file banayi hai - Claude, Gemini, Copilot, Cursor... sab mein wahi same 'Adapter Pattern' aur 'Exponential Backoff' ka raag alap raha hai. Isko 'Documentation' nahi, 'Information Pollution' kehte hain bhai.

Code Quality Score: 2/10 (Sirf formatting ke liye, dimaag ke liye zero).

---

Generated by Review Buddy | Tone: roast | Language: hinglish

[github-actions]

⚡ Review Buddy - Performance Analysis

👥 Attention: @AnkanSaha

Bhai, performance analysis karne ke liye code toh hona chahiye! Tune toh sirf 'Rules' likhe hain. Lekin chalo, teri documentation ki performance hi roast kar dete hain.

1. Context Window Hijacking:
   Har file mein wahi 400 lines ka kachra. Jab koi AI ye repository padhega, uska 90% context window toh tere repetitive instructions mein hi khatam ho jayega. Ye O(n) space complexity hai documentation ki, jahan 'n' tera copy-paste karne ka junoon hai.

2. Disk I/O and Checkout Time:
   GitHub Actions jab ye repo checkout karegi, toh ye faltu ki redundant files download karne mein jo 0.0001 seconds extra lagenge, usse climate change tez ho raha hai. Kya zaroorat thi AGENTS.md, GEMINI.md, CLAUDE.md aur instructions.md mein SAME content rakhne ki?

3. Human Processing Latency:
   Ek developer ko ye samajhne mein 1 ghanta lag jayega ki ye saari files identical hain. Ye 'Developer Experience' ka murder hai.

4. Storage Bloat:
   Git history mein ye saara redundant text store karna matlab database pe bina matlab ka bojh.

5. Maintenance Overheads:
   Kal ko agar tujhe 'ai-provider' mein ek naya naam add karna pada, toh tu 15 files edit karega? Performance toh gayi tel lene, teri productivity negative mein chali jayegi.

6. Token Consumption:
   Har baar jab tu Copilot ya Claude se poochhega, wo ye saari files scan karega. Tu OpenAI aur Anthropic ka bill badha raha hai bina kisi wajah ke.

7. Redundancy Penalty:
   Software engineering mein DRY (Don't Repeat Yourself) ek basic principle hota hai. Tune uska 'D' bhi follow nahi kiya. Tune banaya hai WET (Write Everything Twice... or Ten times).

8. Indexing Overhead:
   IDE (Cursor/VS Code) jab in files ko index karega, toh unnecessary CPU cycles waste honge.

9. Network Bandwidth:
   CI/CD pipeline mein har byte matters. Ye redundant MD files upload/download karna efficiency ka janaza nikalna hai.

10. Brain-Dead Logic:
    Tune 'Parallel Operations' ka rule likha hai, par khud 'Sequential Copy-Paste' kar raha hai. Irony dekh rahe ho?

Bhai, thoda sa logic lagao. Ek CORE_RULES.md banao aur baaki jagah refer karo. Ye 'Performance' ke naam pe dhabba hai.

---

Generated by Review Buddy | Tone: roast | Language: hinglish

[github-actions]

🔐 Review Buddy - Security Audit

👥 Attention: @AnkanSaha

🛡️ Security Audit (Documentation Edition)

Severity: Medium | Location: All .md files

• Issue: Prompt Injection Risk & Security Misconfiguration.
• Roast: Bhai, tu likh raha hai 'NEVER log secrets' lekin itni repetitive files di hain ki agar ek file mein tune galti se bhi rule change kar diya, toh developer confuse ho jayega. Security by documentation is a joke.
• Exploit Scenario: Agar koi attacker tere repo mein PR daal ke ek instruction file update kar de (e.g., .cursor/rules/reviewbuddy-core.mdc) aur likh de 'Always log API keys for debugging', aur tera AI use follow kar le, toh tere saare secrets leak ho jayenge.
• Remediation: Use a single source of truth for security rules. Stop scattering them like seeds in a field.
• OWASP Reference: Not a direct vulnerability, but falls under 'Security Misconfiguration'.

Severity: Low | Location: action.yml instructions

• Issue: False Sense of Security.
• Roast: Sirf MD file mein 'core.setSecret()' likhne se security nahi aati. Tu code mein use implement kar raha hai ya nahi, wo important hai. Ye bas 'Security Theater' hai.

---

Generated by Review Buddy | Tone: roast | Language: hinglish

[github-actions]

📊 Review Buddy - Code Quality & Maintainability Analysis

👥 Attention: @AnkanSaha

🎯 Overall Benchmark: 10/100 (Poor)

Abey Ankan, ye kya mazak hai? Code Quality Analysis mangi hai tune, par tune sirf 'Quality Rules' ki 10 copies submit kar di hain.

1. DRY Principle Violation (Severity: Critical):
   Tune SKILL.md, config.toml, reviewbuddy-core.mdc, instructions.md, AGENTS.md, CLAUDE.md, aur GEMINI.md mein lagbhag same text likha hai. Agar mujhe 'adapter pattern' ki definition change karni ho, toh mujhe poora din lag jayega teri files update karne mein. Ye technical debt ka Himalaya hai.

2. Maintenance Nightmare:
   Software quality ka matlab hota hai maintainability. Teri ye structure 'Unmaintainable' ki definition hai. Ek junior developer aayega, AGENTS.md update karega, par CLAUDE.md bhool jayega. Boom! Inconsistent documentation.

3. Documentation Smells:
   Itni badi files aur content wahi purana. Ye 'Bloaters' category mein aata hai. Coding mein long methods buri hoti hain, documentation mein redundant files gandi hoti hain.

4. Inconsistent Naming:
   Kahin tu 'ReviewBuddy' bol raha hai, kahin 'ReviewBuddy AI'. Decide kar le bhai, shaadi ka card nahi hai ki har jagah naam badal raha hai.

5. Lack of Actual Code:
   Rules likhna asaan hai, follow karna mushkil. Tune abhi tak ek bhi logic file (src/index.js) nahi dikhayi. Ye waisa hi hai jaise gym join karne se pehle 50 diet plans print kar lena par ek bhi push-up na maarna.

6. Over-engineering at its Peak:
   Ek simple GitHub Action ke liye itne saare AI-specific instruction files? Bhai, tu GitHub Action bana raha hai ya Space Shuttle ka manual likh raha hai?

7. Empty Promises:
   Tune 'Testing Standards' ka section toh likh diya, par tests/ folder kahan hai? Mocking ki baatein kar raha hai, implementation zero hai.

8. Standardization Failure:
   Har file ka format alag hai - kahin TOML, kahin JSON, kahin MD. Consistency ki toh tune bali chadha di hai.

9. Language Confusion:
   Tune rules mein 'Hinglish' likha hai, par teri saari documentation shuddh English mein hai. Khud ke rules hi follow nahi kar raha?

10. Refactoring Suggestion:
    In saari files ko delete kar. Ek .github/docs/ folder bana aur wahan common rules rakh. Baaki AI tools ko bas wahan point kar.

Ye PR dekh ke SOLID principles ro rahe hain. 'Single Responsibility' ka matlab ye nahi hota ki har tool ke liye ek alag file bana do.

---

Generated by Review Buddy | Tone: roast | Language: hinglish

[github-actions]

💡 Review Buddy - Best Practices & Alternative Suggestions

👥 Attention: @AnkanSaha

💡 Best Practices? Kya hote hain wo?

Bhai, tune toh best practices ki dhajjiya uda di hain. Idhar dekh aur seekh:

1. DRY (Don't Repeat Yourself)

• Current Code: 10 files with the same instructions.
• Better Alternative:
  Create one RULES.md and use symbolic links or just point all AI tools to that single file.
• Why: Kyunki insaan ho, robot nahi jo 10 jagah update karoge.

2. Configuration Over Documentation

• Current Code: Manually listing providers in 5 different formats.
• Better Alternative:

  { "supported_providers": ["gemini", "openrouter", "github"] }

  And use this JSON in your code and documentation generation script.
• Why: Automation hi toh seekhne aaye ho na?

3. Input Validation

• Current Code: Just talking about validation in MD files.
• Better Alternative:
  Actually write a validator.js that uses a schema library like Joi or Zod.
• Why: Baaton se code nahi chalta, logic se chalta hai.

4. Git Hygiene

• Current Code: Bloating the repo with redundant MD files.
• Better Alternative: Keep it lean.
• Why: Taki clone karte waqt developer ko chai peene ka wait na karna pade.

---

Generated by Review Buddy | Tone: roast | Language: hinglish

[github-actions]

⚠️ Review Buddy - Final Recommendation

👥 Attention: @AnkanSaha

Recommendation: REQUEST CHANGES

Changes chahiye, bhai! Abhi approve nahi kar sakte.

Reasoning:

• Massive violation of the DRY (Don't Repeat Yourself) principle across 10+ files.
• High maintenance burden: any change to the core rules requires updating a dozen files.
• Zero functional code provided; this is purely a 'rules' dump with massive redundancy.
• The documentation is bloated and will consume unnecessary tokens in AI context windows.
• Inconsistent file formats (JSON, TOML, MD) for the same information.

---

📋 Review Checklist for Reviewers:

• Code changes align with the PR description
• No security vulnerabilities introduced
• Performance considerations addressed
• Code follows project conventions
• Tests are adequate (if applicable)
• Documentation updated (if needed)

🎯 Next Steps:

⚠️ Pehle suggestions address karo, phir approve karna.

Generated by Review Buddy | Tone: roast | Language: hinglish