Bezalel
Developer Tools
One MCP endpoint gives every AI agent you run, across Claude Code, Cursor, or Codex CLI, the same shared memory, email inbox, and connectors, so switching frameworks stops meaning starting over.
Toolfree
Curated.
Not crowded.
Describe the job. We break it into steps and match each one to a tool from our verified catalog, with the trade-offs made explicit.
385 tools cataloged · Zero sponsored rankings
Know the tool's name? Search the catalog
New This Week
Recent additions across tools, workflows, and guides.
This row of recent additions scrolls by itself and pauses while you point at it or move focus into it.
Catch a color-contrast failure before it ships, not after a user complains they can't read it.
WorkflowAdvanced · 3 hours, one-time setup
DeepSeek shipped V4.1-Flash with MIT weights and a price cut, announced its V4-Pro flagship would be swapped underneath an unchanged model name, then reversed a day later. The decision for a solo operator is not which model wins. It is whether your stack would have noticed.
ArticleSep 15, 2026 · 7 min read
Instinct
Agents
Text or call a personal AI assistant that actually books the ride, follows up on the dropped email thread, and arranges the handyman, not just suggests it. Currently invite-only.
Toolfree
Give every participant a live meeting in their own language, with no round-trip delay waiting for a full sentence.
WorkflowAdvanced · 3 hours, one-time setup
He tests AI video models by making an actual ad with them and says what each second cost. The last seat on the list, the most perishable, and the one with the most to disclose.
ArticleSep 12, 2026 · 4 min read
Expertise AI
Agents
Install a GTM expert's exact sales playbook as a one-click AI skill that runs on your own CRM and pipeline data, instead of paying a consultant to explain it.
Toolfreemium
Stop paying for the same prompt twice, in about 30 minutes of setup.
WorkflowIntermediate · 30 minutes, one-time setup
Seven years of hands-on tool reviews with a dated pricing table, a 'skip it if' section, and a day job at a company whose product competes with some of what he reviews. He discloses it; we repeat it.
ArticleSep 12, 2026 · 4 min read
Supernova
Developer Tools
Sync your company's tools into one encrypted data lake and query all of it in plain language through Claude or Codex, no data warehouse or data team required.
Toolpaid
Keep your app running when a model provider goes down, in about 3 hours of setup.
WorkflowAdvanced · 3 hours, one-time setup
An AI filmmaking school that tests video models head to head and reads the licence before recommending one. It sells the course, and it sits inside a studio. Both are on the page.
ArticleSep 12, 2026 · 4 min read
Start from a common job
Workflows we recommend
One place to start and two alternatives, with the fit and the trade-offs made explicit.
Start here
Intermediate
Recurring: 30 min per source piece. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Draft platform-specific posts and clips from one article, recording, or webinar.
- For
- Creators with useful long-form material to reuse.
- Watch for
- Edit for each channel; automated reformatting does not preserve your voice on its own.
How the steps connect
Order of steps from this workflow's entry. Each handoff is a review point, not an automatic integration.
Intermediate
One-time setup: 2 hours, one-time setup. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Connect welcome messages, account setup, and progress checks in a repeatable flow.
- For
- Small teams repeating the same onboarding steps.
- Watch for
- Test access permissions and exception handling before letting the flow touch customer accounts.
Intermediate
Estimate: 2-3 hours for a batch of 5-8 interviews. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Group interview transcripts into themes with attributed quotes you can check.
- For
- Founders and researchers reviewing a small interview batch.
- Watch for
- Review every interpretation; a small interview sample is not statistical proof.
Start here
Intermediate
Setup, then recurring: 3 hrs to set up, 30 min per report after. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Produce a recurring report where every figure and citation traces back to a checked source.
- For
- Operators who publish the same report on a schedule and need the numbers to hold up.
- Watch for
- Approve every draft before it goes out; a renamed field or mixed reporting period can turn valid inputs into a misleading narrative.
How the steps connect
Order of steps from this workflow's entry. Each handoff is a review point, not an automatic integration.
Beginner
Recurring: 1.5–2 hrs per article. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Move from a source-backed brief to an edited article draft in about 1.5 to 2 hours, as a planning estimate.
- For
- Solo creators publishing on a schedule without a research or editing team.
- Watch for
- Fact-check every section yourself; a draft can follow the heading while inventing a supporting example.
Advanced
Setup, then recurring: 3–4 hrs setup + weekly review. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Give focus habits and deadline tasks calendar space, then see each week what actually got protected.
- For
- Operators whose priorities get decided by whoever books the next meeting.
- Watch for
- Missing effort estimates produce a tidy calendar you cannot complete; test one synced task before expanding.
Start here
Beginner
Recurring: 2 hours per week. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Leave each week with a reviewed queue of channel-specific posts and notes for the next calendar.
- For
- Small teams keeping a few channels active in about two hours a week after setup.
- Watch for
- Two schedulers can duplicate a LinkedIn post or leave a channel without its asset; reconcile both queues against the calendar.
How the steps connect
Order of steps from this workflow's entry. Each handoff is a review point, not an automatic integration.
Advanced
Estimate: 3–4 hrs for a small prototype; security review and production hardening extra. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Get a reviewed specification, tested endpoint code and a restricted staging deployment in one working session.
- For
- Experienced developers scoping a narrow prototype, not a production service.
- Watch for
- Security review and hardening are extra work; configuration that runs locally can still expose secrets.
Beginner
One-time setup: 1 hour, one-time setup. Planning estimate from the workflow entry, not measured time saved.
- Outcome
- Document your tone, phrasing and examples once, then use the guide as a review checklist for every draft.
- For
- Small teams whose drafts sound like a different writer each time.
- Watch for
- Approve the guide and finished copy yourself; general grammar suggestions can erase deliberate voice choices.
Decision guides
Clearer answers before you choose a tool or build a workflow.
Productivity
Can you trust a directory's ratings?
Know what to check behind star ratings and review counts before they influence a buying decision.
6 min read
Marketing
Keep your voice in automated content
Learn where to add examples and editorial approval so an automated brief does not produce generic copy.
5 min read
Agents
When does a research SOP need an agent?
Separate repeatable instructions from changing inputs before choosing a larger agent framework.
6 min read
Built Around the Job, Not a Tool List
We evaluate tools for functionality, reliability, value, and integration. Our recommendations explain where each tool fits, where it falls short, and what to consider before you commit.
Read our methodologyWhat changed. What matters.
Meaningful AI updates, revised recommendations, and workflows worth revisiting. A weekly brief focused on what changes your next decision.