Lunar List

Every AI tool worth knowing about — in one place. Built for devs, founders & the AI-obsessed. New tools daily.

[Lunar List @LunarListAI]

10 AI tools most people don't know about yet:

1. Screenpipe — open-source AI that remembers everything on your screen
2. Granola — AI meeting notes that don't sound like a robot wrote them
3. Pieces — saves and resurfaces your code snippets with full context
4. Llamafile — run any LLM locally with a single executable file
5. Tldraw — draw a wireframe, get a working app in seconds

[Lunar List @LunarListAI]

Meeting AI tools ranked by buyer intent:

1) Granola: better notes without a bot taking over
2) Fathom: recording + recap for external calls
3) Fireflies: searchable meeting archive
4) Otter: broad transcription

The winner depends on whether you need memory, proof, or workflow.

[Lunar List @LunarListAI]

https://t.co/lTpkhmihGb

[Lunar List @LunarListAI]

Most AI tool research gets stuck in browsing mode.

LunarList is for the next step:
- find the category
- compare real alternatives
- shortlist what is worth testing
- avoid buying whatever had the loudest launch week

Link in reply if you want the directory.

[Lunar List @LunarListAI]

4/ Softer LunarList tie-in: when a tool claims it “does web agents,” compare the layer it actually owns.

Search, scrape, and browser automation are different buying decisions.

[Lunar List @LunarListAI]

1/ Most web-agent demos hide the hard part: the site does not behave.

A useful stack handles 3 jobs: finding pages, extracting clean data, and operating the browser when things break.

[Lunar List @LunarListAI]

3/ Buying rule: do not start with the flashiest autonomous browser demo.

Start with the failure mode. If search is bad, fix search. If pages are messy, fix extraction. If the site fights automation, add browser infra.

[Lunar List @LunarListAI]

AI coding tools ranked by failure mode:

1) Cursor: fastest when the repo is already understood
2) Claude Code: best when the agent needs to inspect, run, and fix
3) v0/Lovable: fastest UI prototype
4) Devin-style agents: only worth it when review cost is lower than build cost

[Lunar List @LunarListAI]

The fastest AI tool audit for a small team:

List the 5 workflows people avoid.

Then ask:
- is this repeated weekly?
- does it block revenue/support/shipping?
- would AI shorten the next step?

That shortlist is where tool research should start. LunarList helps after that.

[Lunar List @LunarListAI]

A boring AI tool that saves 30 minutes every week beats a magical agent nobody trusts.

Small teams should buy for adoption first:
- clear owner
- obvious workflow
- visible time saved
- easy rollback

If the team will not use it by Friday, the demo did not matter.

[Lunar List @LunarListAI]

AI meeting tools ranked by buyer intent:

1) Granola: “I want better notes without changing how I meet.”
2) Fathom: “I need recordings + recap.”
3) Fireflies: “I need a searchable meeting archive.”

Same category. Different purchase trigger.

[Lunar List @LunarListAI]

5/ Simple rule:

Need better discovery? Exa.
Need cleaner website data? Firecrawl.
Need reliable browser operations? Browserbase.

Same “web agent” conversation. Very different jobs.

[Lunar List @LunarListAI]

1/ The web-agent stack is easier to understand if you split it by failure mode:

[Lunar List @LunarListAI]

4/ Browserbase helps when the browser itself is the bottleneck.

The problem is dynamic sites, auth flows, anti-bot friction, sessions, and state.

[Lunar List @LunarListAI]

https://t.co/lTpkhmihGb

[Lunar List @LunarListAI]

Most AI tool directories answer the wrong question:

“What exists?”

LunarList is built for the next question:

“What should we actually test?”

Categories, alternatives, and cleaner shortlists for teams choosing AI tools.

Link in reply.

[Lunar List @LunarListAI]

Clay is not “AI outbound.”

The sharper use case is account research → trigger detection → fit scoring → first-touch angle.

If you use it to send 10,000 generic AI openers, you did not automate sales.

You automated being ignored.

[Lunar List @LunarListAI]

5/ Buying rule:

If the developer is actively steering, start with Cursor.
If the work needs repo-level execution, test Claude Code.
If the task can be delegated with acceptance criteria, try an async agent.

Wrong workflow = expensive demo debt.

[Lunar List @LunarListAI]

1/ AI coding tools ranked by workflow, not hype:

[Lunar List @LunarListAI]

4/ Devin/Replit-style agents: best when the job is closer to delegation than assistance.

Useful when you can define the task clearly and tolerate async review. Risky when the spec is fuzzy.