Now

955 opinionated gems across 13 stacks. Unique titles, markdown pages, /llms.txt, JSON /api/gems, /for-agents, a read-only MCP server at /api/mcp, best-of lists, RSS, and /submit. Daily 7:27 ships a handful of real gems plus GEO, and Monday kills what stopped earning its spot.

Honest catches, not marketing copy:

• →Gemini CLI: consumer cutoff 2026-06-18. Enterprise and API keys stay. Antigravity CLI is the successor, not 1:1.
• →Crush: FSL-1.1-MIT — competing-use restriction, MIT after two years, source-available, not OSI. A TUI that still loads MCP.
• →Chrome DevTools MCP: chrome-devtools-mcp@1.8.0 engines “^20.19.0 || ^22.12.0 || >=23”. 20.16/20.17 fail; 20.19 LTS is in engines. README “LTS” is marketing; engines is the catch. CLI exists; official CLI docs call it experimental.

We list tools with a free path (OSS, free, or a real free tier). Paid-only is an exception with a loud Catch. No sponsors. No affiliate links. Tools earn the spot.

Just Landed on the homepage. The catalog expands daily. Pages are agent-readable so bots can cite and revisit: unique titles, markdown, JSON, RSS, llms.txt.

Measurement: PostHog (EU). Traffic is still a handful of visitors a day — honest number, early days.

Still giving it away. Next, not next quarter's invoice:

• →Comparison pages — honest side-by-sides when the catalog has enough overlap.
• →Monday kill/keep — what stays in the catalog, what doesn't earn another week.
• →A newsletter only once the email list isn't empty. We will not launch a letter into a void.

Not selling rankings. No ads on the catalog. No paid slots until this is actually worth coming back to.

Sequence: lots of free value, then humans and bots visiting and revisiting, then more free value, then — last — a product picked from proven demand. Charging now would be too soon.

When people and agents already use this, we pick a product from that demand. Not before.