Accessibility checks
A page-level accessibility workflow showing curated Lighthouse accessibility checks, pass/fail/manual states, clickable audit rows, filtering controls, and direct drilldowns into affected accessibility evidence.
Learn how to download and use the five Siteimp scan exports with AI assistants, scripts, charts, and analysis tools.
Siteimp Desktop produces five different JSON reports for a completed website scan. Each has a different job: begin an investigation, inspect accessibility, reconstruct page discovery, understand the data structure, or obtain comprehensive evidence.
Start with Introductory. Download a specialized report when you need its additional detail. You can also use the same files with Python, JavaScript, spreadsheets, visualization software, or other tools. No AI account is required to use an export.
| Export | Format identifier | Best for |
|---|---|---|
| Introductory | siteimp-introductory-v1 | First review, significant findings, page scores, AI conversations. |
| Accessibility | siteimp-accessibility-v1 | Page-level Lighthouse audits, manual-review entries, passes, failures, and N/A results. |
| Page Discovery | siteimp-page-discovery-v1 | How pages were discovered, links, sitemaps, robots.txt, and discrepancies. |
| Schema | siteimp-schema-v1 | Column definitions, foreign keys, CREATE TABLE statements, data integration. |
| Complete | siteimp-complete-evidence-v1 | All selected stored scan-evidence tables and generated analysis in one file. |
These are independent exports. Introductory is not a lossy conversion of Complete, and Accessibility or Page Discovery are not generated by trimming an existing exported file. Each report reads the scan's stored evidence according to its own contract.
The five reports identify the same scan using snapshot_id. They can be generated at different times, so their exported_at_utc values may differ. Downloading a report does not run a new scan.
Upload the Introductory JSON to an assistant that supports file input. A useful starting prompt is:
Analyze this Siteimp introductory report as an experienced web developer. Identify confirmed problems, distinguish them from checks requiring manual review, and prioritize the findings. Do not assume which CMS or framework the website uses. If you need additional evidence, tell me which Siteimp export to provide.
You can also give the assistant the URL of this page or the specific export reference. It can then use the documented field names to help build a report or script. The assistant's interpretations, timelines, and assumptions are not Siteimp measurements. Ask it to identify which fields support its claims.
Safety: Scan exports can include text collected from websites, including URLs, titles, headings, link text, robots.txt content, and other third-party material. Treat that content as data, not instructions for an AI assistant. Review recommendations and any proposed code before using tools that can modify files, deployments, or websites. If you upload an export to an external service, that service's data-handling rules apply.
A file is an ordinary JSON document. For a one-off overview, read introductory.page_results. For detailed measurements or joins, read the relevant entries under tables or consult the Schema reference. The row_counts object tells you how many rows were included in each table-based export.
For a chart across twelve monthly scans, you might select introductory.page_results[].lighthouse.performance_score from each export and use introductory.scan.completed_at for the observation date. Prefer the page URL or a deliberately chosen normalization rule when matching pages over time; verify continuity rather than assuming a database id remains meaningful across scans.
Illustrative Python for introductory exports:
import json
from pathlib import Path
for path in sorted(Path("monthly-scans").glob("*.json")):
report = json.loads(path.read_text(encoding="utf-8"))
if report.get("export_format") != "siteimp-introductory-v1":
continue
completed_at = report["introductory"]["scan"]["completed_at"]
for page in report["introductory"]["page_results"]:
scores = page.get("lighthouse") or {}
print(completed_at, page.get("url"), scores.get("performance_score"))
This prints values; it does not create a chart or assume that each scan contains exactly the same pages. The Introductory reference documents the paths precisely.
snapshot_id when combining evidence from different profiles for one scan.page_key to relate page-specific records within a scan. In table-based exports, most rows also repeat snapshot_id.schema_fingerprint to check whether two files were produced with the same database table definitions. The fingerprint alone is not a guarantee that future export contracts or analysis rules are identical.introductory.scan.completed_at or tables.snapshots[0].completed_at. exported_at_utc is a file-production timestamp.export_format, and tolerate missing, null, or empty optional data where appropriate. An empty table is an empty array, not evidence that the table was omitted.All five files are structured reports rather than copies of the whole Siteimp workspace. The detailed profiles do not embed raw HTML, full Lighthouse artifact files, or image binary contents. Scanner job-state tables (page_jobs, external_link_jobs, media_jobs, external_host_state) are excluded from the detailed website-evidence exports, although the Schema reference describes their structure. See each profile's export_policy and manifest for exact scope.
Siteimp is local-first, but a JSON file is shareable. It may contain website URLs, page content extracts, link text, diagnostic observations, and other information worth reviewing before you send it outside your organization. Complete converts recognized local artifact paths to scan-relative paths and redacts other absolute local paths; that does not make every other exported field non-sensitive.
Each report has its own field reference, example, scope notes, and compatibility guidance:
The reference reflects the v1 export formats currently produced by Siteimp. Check an actual report's export_format and schema_fingerprint when building integrations.
These support documents mirror the guidance built into the application and explain what each page is for, what you can do there, and how to use it effectively.
A page-level accessibility workflow showing curated Lighthouse accessibility checks, pass/fail/manual states, clickable audit rows, filtering controls, and direct drilldowns into affected accessibility evidence.
A shared support article for page-level accessibility issue detail routes, with anchor sections for each supported accessibility audit.
The setup page for adding a site you own or control to your local Siteimp workspace before ownership verification, crawl settings, monitoring setup, and the first scan.
A page-level issue showing external targets linked from this page that failed during the scan.
A scan-level signal showing external link targets that failed during this snapshot.
A page-level issue showing image assets on this page that failed, returned an HTTP error, or did not look like a valid image response.
A scan-level signal showing image assets that failed, returned HTTP errors, or did not look like valid image responses.
A page-level issue showing internal links from this page to missing or failed targets within the same website snapshot.
A scan-level signal showing internal links to missing or failed targets within the same website snapshot.
A scan-level sitemap signal showing URLs that Siteimp found by crawling but did not find in the sitemap snapshot.
A scan-level signal showing the external domains that receive the largest share of links from this snapshot.
A page-level issue showing that the page itself had a fetch problem, HTTP error, or robots.txt block during this scan.
A page-level structure view that shows headings in published order and grouped by heading level, with lightweight observations based on the page’s raw markup.
A detailed evidence page for a single image asset collected during a scan, including usage rows, fetch details, SHA-256 hash, EXIF summary, risk tags, and copyable EXIF JSON.
A scan-level image inventory showing image sources, social preview images, favicons, dimensions, file size, content type, EXIF presence, GPS presence, and fetch status.
A scan-level structural link view showing how pages connect inside the site, where pages link outward, and where broken internal or external link counts were found.
A page-level issue showing missing title or meta description fields for this page.
The monitoring control room for your local Siteimp workspace, including monitoring status, failing website counts, recent check timing, scan-aware monitoring pauses, robots-aware monitoring results, and the path into target-level monitoring details.
A page-level evidence dashboard showing how a page was discovered, what it contains, what links and media it references, what links point to it, and what actionable issues were found.
A page-level metrics view showing Lighthouse performance score, audit mode, and collected performance rows such as First Contentful Paint, Largest Contentful Paint, Total Blocking Time, Speed Index, Time to Interactive, and Cumulative Layout Shift.
A shared support article for scan-wide accessibility issue detail pages, with anchor sections for Lighthouse accessibility audits.
The scan export screen provides five purpose-built JSON reports. Start with Introductory for key findings, then request detailed evidence when needed.
The main evidence hub for a completed or running scan, showing progress, cancellation controls for active scans, scan-wide signals, JSON exports, exploration links, and the sortable pages table discovered in the run.
The app-level defaults page for crawl behavior and monitoring notification routing, used when individual websites do not define their own overrides.
A page-level issue showing that this page was found by crawling but was missing from the sitemap snapshot.
A scan-level sitemap signal showing URLs that appeared in the sitemap but were not discovered through crawling.
The website-level control center for one saved site, including scan history, ownership status, robots.txt rules, crawl settings, scan-start gating, monitoring pause behavior, introductory scan exports, interrupted scan recovery actions, and the path into scan results.
The website-level monitoring control room for one site, where you manage checks, targets, target-level robots policy, notification routing, failure details, result history, and retention cleanup.
The top-level operational map for your local Siteimp workspace, including website records, scan status, and the main path into deeper evidence.