Port of the TypeScript compiler, checker and lsp to Rust, by LLM

Tracks actual LLM API token consumption versus subscription limits and provides real‑time USD cost estimates to eliminate confusion over “API priced tokens.”
Core value: prevents surprise spend and lets developers understand the true financial impact of LLM experiments.

Key	Value
Target Audience	Developers, researchers, AI hobbyists using LLM APIs via subscriptions (Claude, OpenAI, etc.) who want to monitor spend
Core Feature	Dashboard that logs token usage from API keys, converts to API‑equivalent cost, compares to the user’s subscription plan, and sends alerts when approaching limits
Tech Stack	Frontend: React (TypeScript); Backend: Node.js/FastAPI; Database: PostgreSQL; Optional: lightweight proxy/middleware to intercept LLM API calls
Difficulty	Medium
Monetization	Revenue-ready: SaaS subscription tiered by number of monitored accounts (e.g., $5/mo per user)

HN commenters expressed frustration over opaque cost accounting: orthogonal_cube questioned “They spent over $400k for an experimental project …”, while flossly noted “Total token spend was ~$24,047 of API spend over 2 weeks.” A clear tracker would address this pain.
Potential to spark discussion about LLM economics, help teams avoid wasteful spending, and integrate into CI pipelines for cost governance.

Automatically validates equivalence between original and LLM‑generated code using test suites, property‑based testing, and symbolic execution, emitting diagnostic messages on mismatches.
Core value: increases trust in LLM porting efforts by catching regressions early and reducing manual verification burden.

Key	Value
Target Audience	Engineers porting codebases with LLMs, open‑source maintainers, companies performing AI‑assisted migrations
Core Feature	Given source and target repos, runs the existing test suite on both, generates additional fuzz tests, and reports semantic divergences with concrete counterexamples
Tech Stack	Python; pytest & hypothesis for property‑based testing; Z3 or K Framework for lightweight symbolic checks; Docker for isolation; GitHub Actions CI integration
Difficulty	High
Monetization	Revenue-ready: per‑project pricing for enterprise verification (e.g., $0.10 per 1k LOC verified)

hedgehog asked for “as much as possible mechanical (non‑LMM) verification of equivalence that generates good diagnostic messages,” and anonymous908213 stressed the value of feeding LLMs years of human test suites. VeriPort directly satisfies these requests.
Could become a standard CI step for LLM‑generated migrations, fostering discussion on reliability vs. speed trade‑offs and encouraging better test coverage.

Combines rule‑based source‑to‑source translation with selective LLM assistance for ambiguous constructs, cutting token usage and improving port quality.
Core value: lowers cost and increases reliability of language migrations (e.g., TypeScript → Rust, Go → Rust) by automating the straightforward parts and reserving LLMs for tricky sections.

Key	Value
Target Audience	Developers performing large‑scale language migrations, tooling teams, language maintainers
Core Feature	Parser (tree‑sitter) builds AST; deterministic rewrite rules handle clear cases; unresolved nodes invoke a focused LLM prompt; results are cached and combined into idiomatic target code
Tech Stack	Rust (performance); tree‑sitter for parsing; Serde for AST manipulation; optional LLM API (OpenAI/Anthropic); WASM playground for demo
Difficulty	Medium‑High
Monetization	Hobby (open‑source) – MIT license, with optional sponsored extensions for enterprise support

spankalee suggested “investing in really good deterministic automatic translators that do 90% of the job cheaply and delegate any decisions that have to be made back out to the AI agent?” – exactly what Translattice provides.
dwattttt found the solo Go porting effort “interesting to read about”; a hybrid tool would reduce the token burden and make such experiments accessible to more people, inviting discussion on the balance between automation and LLM creativity.

Provides a web‑based environment where LLMs can compile language frontends (e.g., TypeScript) to WebAssembly, with size and performance benchmarks, enabling immediate in‑browser type checking and experimentation.
Core value: lets developers evaluate LLM‑ported compilers in WASM without local setup, focusing on performance, size, and editor integration.

Key	Value
Target Audience	Frontend engineers, language‑tooling researchers, educators
Core Feature	One‑click build of a compiler to WASM (via LLVM/Emscripten or Rust/wasm‑bindgen), runs in a Web Worker, shows compile time/memory usage, and offers an API for IDE integration
Tech Stack	Rust/C++ compiled to WASM (wasm‑pack/emscripten); Frontend: Svelte or React; Web Workers; IndexedDB for caching; optional Monaco Editor for demo
Difficulty	High
Monetization	Hobby (open‑source) – MIT license, with possible paid tiers for private projects and advanced benchmarking

Benjamin_Dobell noted the lack of WASM support for the TypeScript checker stucks them on TS 6, while viraptor described Theo’s vision of an IDE that hosts compilation and type‑checking in WASM. Wasmify directly enables that experimentation.
Could drive conversation about WASM adoption for language tooling, encourage benchmark sharing, and lower the barrier for trying LLM‑generated compilers in the browser.

🚀 Project Ideas