Hayz Labs · https://hayzlabs.com/tools/
22 tools · they work right on the page
Hands-on tools
Small interactive tools: sort examples, put in your own numbers, check your answer. Each shows the rule it used, so you can disagree with the rule instead of the answer. They work right on the page and save nothing.
- Entries
- 2
- Last change
- 10 Oct 2026
- Beginner
- 1
- Intermediate
- 1
Am I actually faster?Median minutes with and without AI, including review and fixes, from your own runs, with a quality check beside the time.Intermediatev1.110 Oct 2026Where could AI help this week?Score up to 15 recurring tasks by minutes a week, how repeatable each is and how easily you can check the result. The three highest scores are the tasks to try with AI first.Beginnerv1.110 Oct 2026
No tools match.
Inside other pages
These run on the page they belong to, next to the steps they support.
What not to pasteA three-tier rule and a two-minute settings check, so nothing you regret goes into a chat.balanceAI detectors: what they can and can't tell youA detector flag is a guess about a text and no proof about a person; independent tests found false positives and unreliable results, and a 2025 paper found some tools much better, so a flag is a reason to ask.prioritiesAI voice scams: hang up and call back on a number you knowA call-back rule on a number you already have, a short card for older relatives, and an optional family phrase as an extra check, with the official 2025 advice behind it.prioritiesAI-written phishing: stop looking for typos, set up passkeysScam emails no longer give themselves away with bad spelling. A 20-minute setup that protects your accounts even when you cannot tell a fake from the real thing.outputBuild a small agent yourself: one tool, a step limit and a logAn agent is a model calling tools in a loop. Build the smallest useful one: a read-only search over a folder of notes, a hard limit of six steps, a token and cost tally, a log of every tool call, and answers that must cite their files. Then test it on nine questions, one aimed at a planted instruction.outputCheck before you trust: a five-minute routineA seven-line checklist you run on any AI answer before you act on it.outputDo I need to pay for AI, and for which plan?When a free plan is enough, when one paid plan is worth about $20 a month, how to pick that one plan, and a two to four week test before and after you pay.prioritiesFind near-duplicates and group a long list by meaning (embeddings)Turn each item in a long list into numbers with an embeddings model, list the closest pairs as near-duplicate candidates, group the rest with a short numpy script, and read samples yourself before you name or count anything.outputGet a spreadsheet formula from AI, then prove it worksDescribe the sheet exactly, ask for the formula with its edge cases, run it on 8 to 12 rows you worked out by hand, check for silent failures, and only then fill down and reconcile one known total.outputGet data you can trust from an API: schemas, validation and retriesValid JSON, JSON that matches your schema and JSON with the right values are three different things. Let the vendor guarantee the shape, check the values in your own code, retry once with the errors, then hand the item to a person.outputHealth, money and legal questions: green lights and red linesWhat an assistant can safely do with a health, money or legal question, where you stop, who to ask instead, and how to ask for a person to review a decision software may have made about you.balanceLocal or cloud: when a model on your own device is enoughDecide by what the data is and what the task needs, and check by hand whether a model will fit in your computer's memory, using a rule of thumb of about 0.6 GB per billion parameters.balanceLong agent jobs: pieces, milestones and stop rulesA routine for cutting a big job into checkable pieces for an agent, with a milestone file for each piece, a spend and time cap, a log, and a stop rule that tells you when to take the work back.outputPrompt, retrieve or fine-tune? Getting a model to use your materialName the problem first. Missing knowledge is fixed by putting your material in front of the model. Wrong format or style is fixed by examples, and only rarely by fine-tuning. Climb one rung at a time and measure each rung on the same 20 questions.outputPrompting in one pageThe few prompt habits that vendor docs and published tests agree on, the tricks that make no reliable difference, and how to save a prompt you reuse and change it without breaking it.outputReconcile two exportsTwo files that should agree, such as a bank export and your bookkeeping: a small script matches them by rules you approve and proves nothing was lost, you check a sample by hand, and an assistant only helps label the leftovers.outputRun one prompt over hundreds of rows, then measure how often it's wrongLabel, sort or extract hundreds of rows with one fixed prompt: hand-check a test set first, run in numbered chunks, check the counts mechanically, then hand-check a random sample and turn the errors you find into an error rate with a range.outputScript, workflow or agent: how much to automateEight rungs from doing it by hand to an unattended agent, five questions to pick the lowest rung that does the job, and the tasks where a plain script beats a model.outputTest any new model on your own tasks (a personal eval set)Keep ten to twelve of your own tasks with written pass rules, run them on any new model or plan, judge them yourself, and switch only by a rule you wrote in advance.prioritiesYour first API call, with a spending capSet a hard monthly cap on a separate project before you create a key, keep the key in an environment variable, make one call with curl and Python, and turn the usage fields into a cost. Anthropic and OpenAI side by side, Gemini in one line.output