Missing export path
link.missingindex.html, line 7:7
<a href="visit/">
Read the visitor guide
</a>The export has no visit/index.html. The intended page is guide/index.html, so the link points to the wrong directory.
A worked example / LaunchLens analyzer v1
Find a missing page.
Fix a broken anchor.
A two-page fictional garden site has two deliberate link mistakes. Run the broken export, follow the report to the exact lines, and compare the repaired version.
Run the broken sample →The bundled five-file fixture runs locally in the checker. No account, uploads, or donation required.
2 pagesfive small exported files
2 errorsone path, one anchor
0 findingsafter the two repairs
1. Match the export to its intended URLs.
The sample is called Pocket Garden. It is purpose-built fictional input, not a live venue. Its ZIP contains:
pocket-garden/
index.html
guide/index.html
assets/leaf.svg
robots.txt
sitemap.xmlUse https://pocket-garden.example/ as the deployment base URL. Choose Directory indexes and leave Remove one common enclosing folder checked. The sample button sets these for you.
The outer pocket-garden/ folder is packaging. After it is removed, index.html represents the root URL and guide/index.html represents /guide/. If you leave the outer folder in place, you will test a different path layout and can get misleading missing-file results.
A base URL tells LaunchLens how to resolve local references. The checker does not visit that URL. The reserved .example hostname is intentional; no real website is needed.
2. Read two concrete findings.
The actual broken fixture produces 2 errors, 0 warnings, and 0 notes. It contains 2 HTML pages, 5 files, and 9 checked references. The nine references include links, canonical URLs, the image, and robots/sitemap references, so they are not nine unique destinations.
Missing target anchor
anchor.missingindex.html, line 8:7
<a href="guide/#times">
Check the opening hours
</a>The page exists, but its heading uses id="hours". There is no times ID or named anchor. IDs and anchors are case-sensitive.
The code snippets above are reformatted for reading; report locations refer to the original files inside the ZIP. Open those files as text if you want to inspect the exact lines.
A missing export target is not proof of a live HTTP 404. Your real host might supply redirects or rewrites that are not in the ZIP. Check production behavior separately.
3. Change two links, then check again.
In the sample’s index.html, change href="visit/" to href="guide/", and change href="guide/#times" to href="guide/#hours". No other file needs to change.
<a href="guide/">Read the visitor guide</a>
<a href="guide/#hours">Check the opening hours</a>Choose the repaired ZIP in LaunchLens, keep the same hosting rules, and click Check this export. The bundled clean sample uses those same two repairs.
| Result | Broken export | Clean export |
|---|---|---|
| HTML pages / files | 2 / 5 | 2 / 5 |
| Checked references | 9 | 9 |
| Errors | 2 | 0 |
| Warnings / notes | 0 / 0 | 0 / 0 |
Both downloadable ZIPs were unpacked and analyzed with the same engine as the app. Their reports match these counts. Run timestamps will differ when you repeat the check.
What a clean result means
“Zero findings” means this fixture passes the checker’s bounded static-export checks with these settings. It does not establish live availability, search indexing, accessibility, or that every browser interaction works. After publishing a real site, open its important URLs and test those behaviors too.
Small files you can reproduce.
Each ZIP contains only the original toy HTML pages, an SVG leaf, robots.txt, and a sitemap. Compare the two index.html files to see the repairs. The reports capture the analyzer’s actual output.
Before the repair
After the repair
The clean CSV contains column headers and no finding rows. Reports include local file paths and configured URLs; review reports before sharing your own.
Privacy, coverage, and exact limits.
LaunchLens parses selected files as inert data in a cancelable browser worker. It does not execute their scripts, render their HTML, or fetch their URLs. It has no upload endpoint, accounts, analytics, external fonts, or runtime CDN. Hosting still serves app assets and may retain ordinary access logs.
Files and reports stay in this tab’s memory. Clear files releases the app’s references; closing the tab ends the session. Downloads remain in your browser’s chosen folder. Only select files you intend to inspect, and remove secrets from exports before sharing them with anyone.
- ZIP: at most 25 MiB. Unpacked total: 64 MiB. At most 2,000 files and 500 HTML pages per run.
- At most 16 MiB per file; 2 MiB per HTML, XML, or text file. Text must be UTF-8.
- At most 50,000 URL references and 10,000 findings. Repeated findings with the same code, file, and target are grouped with an occurrence count.
- Stored or deflate ZIPs only. Encrypted, ZIP64, multi-disk, symlink, traversal, duplicate-path, and over-100:1 compression-ratio archives are rejected.
- No live HTTP status, redirects, server headers, JavaScript rendering, rankings, or accessibility audit. No CSS URLs, JS imports, srcset, template-content, form-action, or external-destination checks.
- robots.txt, sitemap, canonical, and hreflang checks are basic signals. There is no full robots group evaluation, XML schema validation, or language registry validation.
Start with the complete built export and the correct deployment rules. If a result looks wrong, check those two assumptions before changing a working link.