Read the payload. Keep the identifiers.
Inspect syntax and structure before moving a payload into an application. A valid JSON number can still lose precision when JavaScript reads it.
A practical desk for developer decisions
Developer tools.
Clear, checkable results.
Format JSON, size CSS, inspect email payloads and estimate video delivery. Six free tools, with the assumptions and examples beside the result.
Find the right tool →Free to use. No account required. Tool inputs are processed in your browser.
Start with the task.
Choose an inspector for data you already have, or a planner for assumptions you want to compare.
Convert the scale, not just the number.
Translate a design specification into relative CSS sizes using an explicit root font size. Compare roots before deciding which rem value to use.
Compare the bill behind the headline rate.
Model sending volume, included allowance, overage and fixed charges using your own rates. Apply the same usage assumptions to every candidate.
Find out why the attachment grew.
Inspect the MIME text you provide and distinguish the size of a file from the size of the email sent over the wire. Encoding, headers and boundaries matter.
Turn bitrate into a checkable budget.
Estimate transfer from average delivered bitrate and watched time, then apply your own transfer rate. Keep bits and bytes distinct and state whether your quote uses GB or GiB.
Separate an advertised peak from an average.
Read bitrate attributes in a pasted HLS master playlist. Compare rendition declarations without fetching streams or mistaking a manifest value for observed viewer traffic.
1. Use a small example first
Reproduce an expected result before trusting a larger job. Avoid pasting credentials, personal data or confidential messages.
2. Keep assumptions visible
Specify root font size, included allowance, average bitrate or encoding. A precise answer to the wrong assumption is still the wrong decision.
3. Verify at the boundary
Inspect the actual API response, serialized email, computed CSS or delivered media before relying on an estimate in production.
The reasoning belongs beside the tool
Six calculations worth checking.
Each example explains a failure mode, shows the input and names the point where a calculator stops being evidence.
01 / JSON formatter
A customer ID is not a quantity
Input / assumption
9007199254740993Result / decision
Keep it quotedJavaScript safely represents integers only through 9007199254740991. Preserve an identifier that arrives as a string. Formatting does not validate an API schema or prove that the data is correct.
Start with a synthetic payload. Fix the source rather than silently changing an identifier.
Read the method and limitations →02 / PX to REM
A 24px heading at two roots
Input / assumption
24 / 16 = 1.5remResult / decision
24 / 20 = 1.2remThe divisor is the computed font size of the root element, not the nearest parent. A 16px browser default is a starting assumption, not a guarantee about a particular stylesheet or user.
Check the computed root in developer tools, then test zoom and user font preferences.
Read the method and limitations →03 / Email cost planner
A lower overage rate can still cost more
Input / assumption
100,000 sends; 20,000 includedResult / decision
$26 at $10 + $0.20/1k80,000 billable sends cost $16 at $0.20 per thousand; adding a $10 fixed fee gives $26. A $20 plan with $0.10 per thousand gives $28 for the same allowance. These are synthetic rates, not provider quotes.
Add taxes, dedicated IPs, support and other fees separately. A cost calculation cannot establish deliverability.
Read the method and limitations →04 / Email payload inspector
A 5 MiB file is not a 5 MiB email
Input / assumption
5,242,880 source bytesResult / decision
6,990,508 base64 charactersBase64 uses 4 x ceil(bytes / 3) characters. At 76 characters per line, CRLF separators add roughly another 184,000 bytes, before MIME headers and boundaries. The serialized message is the final measurement.
Use a synthetic message or remove sensitive headers and content. This inspector does not send email or test recipient limits.
Read the method and limitations →05 / Video delivery planner
Plan a thousand complete views
Input / assumption
5 Mb/s x 600 seconds / 8Result / decision
375 GB for 1,000 viewsA ten-minute view at a constant 5 megabits per second transfers 375 megabytes in decimal units. A thousand complete views transfer 375 GB. At a synthetic $0.08 per GB, transfer alone costs $30.
This excludes overhead, retries, storage and transcoding. Partial viewing and adaptive quality change the delivered average.
Read the method and limitations →06 / HLS manifest inspector
Budget from the right assumption
Input / assumption
BANDWIDTH=6000000Result / decision
AVERAGE-BANDWIDTH=4000000For a ten-minute view, a constant 4 Mb/s assumption implies 300 MB; a constant 6 Mb/s assumption implies 450 MB. That 50% difference illustrates a planning choice. Neither number measures what a player downloaded.
BANDWIDTH describes a peak segment bitrate constraint; AVERAGE-BANDWIDTH describes an average. Verify declarations against encoded media.
Read the method and limitations →Clicked a suspicious link?
That calls for a response workflow rather than a formatter. Clear Response separates opening a page, entering a password, downloading a file and sending money, with next steps and an incident-note template.
Open the security response planner →Know how this site works.
Read how examples are checked, where sources come from, what AI assistance means here and what these tools cannot establish.
Read the publishing standards →