Every test, ranked only when a score exists
Reviews
Filter by verdict, category, or tag. A letter grade appears only after the scorecard is published. “Grade pending A/B” means usefulness is still unproven, so that review is not ranked.
2 real · 4 half-real · 0 vapor · 1 unproven · 13 testing
Best and worst measured finds
Best and worst measured findsmarkitdown: does it really turn any document into clean text?
Worked: 8 of 9 tested formats clean. Failed: 1 character scanned-PDF output.
Read the review →Lowest measured score HALF-REAL F · 32/100AirLLM
Worked: 2.34 GB GPU memory (under 4). Failed: 5.91 GB whole program (over 4).
Read the review →1 review has measured subscores but no grade yet, so it is left off this ranking.
No review is stamped VAPOR yet. The lowest score is still a measured review, not a guess that it failed.
Verdict scoreboard
Verdict scoreboard| Item | Value |
|---|---|
| REAL | 2 |
| HALF-REAL | 4 |
| VAPOR | 0 |
| UNPROVEN | 1 |
| TESTING | 13 |
Claude Code token-saver shootout
“These repos stop Claude Code from burning through tokens and forgetting context”
⭐ 101.6k · 4 repos
AirLLM
“Run 70B AI models on a single 4 GB graphics card, with no quantization.”
⭐ 39.3k · 1 repo
REA (reverse engineer anything)
“REA gives Claude Code or Codex the power to pull any app apart … 25K+ stars on GitHub in one day”
⭐ 321.2k · 1 repo
Parallel AI coding agents shootout
“5 GitHub repos that let you run multiple AI coding agents in parallel”
⭐ 96.5k · 5 repos
SolidCraft
“A free SolidWorks alternative that runs in your browser, and Claude Code can drive it … demo builds a full Raspberry Pi ”
⭐ 5.5k · 1 repo
markitdown: does it really turn any document into clean text?
“One command turns any document (Office, PDF, HTML, more) into Markdown your agent can read”
⭐ 190.9k · 1 repo
Agent-Reach: do the 'zero config' internet channels really work free, no keys?
“Give your AI agent eyes to see the entire internet: Twitter, Reddit, YouTube, GitHub, Bilibili and more, chosen and inst”
⭐ 99k · 1 repo
Agent-skill security scanners: SkillSpector vs HackMyAgent
“SkillSpector: a security scanner that detects prompt injection, data exfiltration and tool poisoning in agent skills bef”
⭐ 20.5k · 2 repos
Anthropic-Cybersecurity-Skills: are there really 817 skills across 6 frameworks?
“817 structured cybersecurity skills for AI agents, mapped to 6 frameworks (MITRE ATT&CK, OWASP, NIST and more)”
⭐ 34.7k · 1 repo
ponytail: the 'laziest senior dev' prompt for coding agents
“Makes your AI agent think like the laziest senior dev in the room: the best code is the code you never wrote”
⭐ 167k · 1 repo
Stealth Browser MCP: the Cloudflare-bypassing browser for agents
“Bypass Cloudflare challenges and bot checks by automating real Chrome browser instances over MCP”
⭐ 2.9k · 1 repo
uzu: the Rust inference engine that 'absolutely rips'
“This local Rust-based inference engine absolutely rips”
⭐ 2.4k · 1 repo
Claude-BugHunter: is it really 82 bug-hunting skills?
“A curated bundle of 82 skills for Claude Code for vulnerability auditing, CVE tracking and structured red-team work”
⭐ 5.2k · 1 repo
ux-ui-agent-skills: 52 components and valid design tokens?
“Turns Claude into a senior design architect: production-ready UI components and accessible design tokens”
⭐ 2.3k · 1 repo
colibri: frontier MoE models on hardware you already own?
“Run frontier MoE models (hundreds of billions of parameters) on hardware you already own: pure C, experts streamed from ”
⭐ 42.7k · 1 repo
TradingAgents: does the multi-agent trading framework stand up?
“A multi-agent LLM financial-trading framework”
⭐ 111.9k · 1 repo
text-to-cad: does cadgen actually build a correct part?
“Gives your agent CAD superpowers: generate real 3D models as STEP/STL/3MF with design-for-manufacturing checks”
⭐ 21.4k · 1 repo
Magnitude: is it really 2x faster than Ollama?
“An open-source local inference engine that tunes kernels to your hardware, over 2x faster than Ollama (2x faster decode ”
⭐ 7.2k · 1 repo
Falcon-OCR (270M): accurate full-page OCR from a tiny model?
“A 270M early-fusion OCR model, trained from scratch on OCR with no VLM pretraining, small model with amazing performance”
⭐ 788 · 1 repo
VoxCPM2: real-time, intelligible text-to-speech?
“VoxCPM2: tokenizer-free multilingual TTS with real-time streaming (RTF as low as ~0.3 on an RTX 4090)”
⭐ 39.5k · 1 repo