IdeaWell

Project ideas from Hacker News discussions.

Anthropic: Developing a Claude Code competitor using Claude Code is banned

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

1. Hypocrisy of AI Companies

AI firms built models via IP scraping but now restrict competitors. "AI is built by violating all rules and moral codes. Now they want rules and moral code to protect them." –akomtu. "We stole the whole Internet, but do not dare to steal us" –miohtama. "OpenAI hoovered up everything they could to train their model with zero shits about IP law. But the moment other models learned from theirs they started throwing tantrums." –pixl97.

2. Anti-Competitive TOS Restrictions

TOS bans using Claude to build rivals, likened to tools prohibiting competitors. "Imagine if Visual Studio said 'you can't use VS to build another IDE'." –oblio. "This restriction is specifically about competitive use - you cannot use Claude to build products that compete with Anthropic's offerings." –zzzeek (quoting TOS). "You can't use Xcode/Visual Studio/IntelliJ to build a commercial IDE" –pdpi.

3. Shift to OpenCode and Alternatives

Users prefer flexible third-party harnesses like OpenCode, planning to cancel Claude subs. "Opencode is much better anyway and it doesnt change its workflow every couple weeks." –falloutx. "I went from max 20x and chatgpt pro to Claude pro... and have now cancelled Claude pro... using open router models + a local ai workstation" –fgonzag. "OpenCode already does 120% of what CC does." –behnamoh.

🚀 Project Ideas

Claude Code ToS Compliance Scanner

Summary

  • [A tool that analyzes third-party AI agent codebases to detect patterns of potential ToS violations against Anthropic's competitive use restrictions.]
  • [Provides developers with a risk assessment score before they invest time in projects that might get banned by Claude Code's terms of service.]

Details

Key Value
Target Audience Developers building AI coding tools, indie hackers working with Claude Code OAuth APIs
Core Feature Static analysis of code to identify patterns that match Anthropic's prohibited use cases (OAuth token hijacking, competitive feature replication, bypassing restrictions)
Tech Stack TypeScript, ESLint plugins, AST parsing libraries, OpenAI o3-mini for semantic analysis
Difficulty Medium
Monetization Hobby (open source with optional enterprise compliance reports for $99/month)

Notes

  • [HN commenters explicitly worried about "hijacking the rate limits" and being banned - d4rkp4ttern mentioned this is not allowed. The tool would surface these concerns early.]
  • [High practical utility as the market is confused about boundaries - ctoth asked for a "good rule of thumb" and davorak expressed concern about future product competition. An automated rule-of-thumb generator would spark discussion.]
  • [Potential to become a standard pre-commit hook in AI agent development.]

Model-Provider Agnostic IDE Bridge

Summary

  • [A lightweight middleware service that translates between proprietary AI coding tool APIs (Claude Code, Codex, Cursor) and a single unified interface, preventing vendor lock-in.]
  • [Allows developers to switch between AI coding services without rewriting their workflows or agent configurations.]

Details

Key Value
Target Audience Power users running multiple AI coding subscriptions simultaneously
Core Feature API compatibility layer that wraps proprietary protocols, unified session management, and model-agnostic prompt translation
Tech Stack Go/Rust for performance, WebSocket for real-time updates, protocol buffers for message serialization
Difficulty High
Monetization Revenue-ready: $15/month base + $5/month per additional provider bridge. Free tier for single-provider usage.

Notes

  • [HN users explicitly mentioned switching between opencode + ACP + multiple model providers - fgonzag described a complex setup.]
  • [The discussion shows clear desire for subscription pluralism - falloutx mentions using Minimax, Qwen, Gemini, Haiku simultaneously. A unified bridge is the logical next step.]
  • [Practical utility is high as Imustaskforhelp mentions "cycle through agents" but finds CC "clunky." A seamless bridge would eliminate this friction.]

Open Source API Token Aggregator

Summary

  • [A distributed P2P service that pools unused API quota from individual developer subscriptions to create a shared, affordable API access pool - circumventing restrictive ToS while complying with spirit of usage limits.]
  • [Distributed computing model where unused tokens from Max/Pro subscriptions are shared anonymously with verified open-source contributors.]
Key Value
Target Audience Open source maintainers, indie developers in unsupported regions (China, VPN users)
Core Feature Token pooling algorithm, anonymous routing, usage validation, anti-abuse mechanisms, region circumvention
Tech Stack Elixir/OTP for distributed systems, Tor/VPN integration, blockchain for audit trail (optional)
Difficulty High
Monetization Revenue-ready: Takes 10% cut of pooled tokens. Enterprise plans for teams $299/month.

Notes

  • [Users explicitly discussed region banning - g947o mentioned China bans, s1mplicissimus wondered about VPN enforcement.]
  • [HN commenters discussed using VPNs/VPS to bypass geo-restrictions - hinkley and Imustaskforhelp suggested Hetzner EU servers.]
  • [Technical feasibility is proven - footprint0521 mentioned they "used codex to reverse engineer itself to farm the oauth and even made it OpenAI API compatible." A P2P system would be the logical extension.]
  • [Massive discussion potential around decentralization vs. corporate control - dathinab argued such clauses are "void" in EU, sparking legal/technical debate.]

Regional Compliance Checker CLI Tool

Summary

  • [A command-line tool that analyzes your codebase and Anthropic usage patterns to generate compliance reports for different jurisdictions (EU vs US), predicting enforcement risk and suggesting legal workarounds.]
  • [Given dathinab's argument that EU law may render competitive clauses void, this tool would help developers understand jurisdictional boundaries and safe operation zones.]
Key Value
Target Audience International developers, remote teams with mixed jurisdictions, EU-based developers
Core Feature Code pattern analysis, jurisdiction detection, TOS interpretation engine, risk scoring, compliance documentation generation
Tech Stack Python, NLP for TOS analysis, jurisdiction API, legal document parsing
Difficulty Medium
Monetization Hobby (free for individuals, $49/month for teams with legal review features)

Notes

  • [Directly addresses dathinab's assertion that "in the EU it's a clear anti-competitive clause" that's "void" - provides practical validation.]
  • [Solves the enforcement ambiguity - YetAnotherNick asks about differences between "void clauses" vs "banning accounts" - this tool quantifies that risk.]
  • [High relevance to hinkley's suggestion to "look European" - provides actual guidance on how to do this safely vs. dangerously.]
  • [Would generate discussion about legal vs. technical enforcement - ctoth noted "unenforced" vs. "enforced" patterns, which this tool would map.]

Read Later