auditmysite

Accessibility audits for real rendered pages, built for CI and modern frontend stacks

Overview

auditmysite is a Rust CLI that audits accessibility against fully rendered pages in Chrome. Instead of scanning raw HTML only, it uses Chrome DevTools Protocol (CDP) and the browser's native Accessibility Tree, so it can evaluate dynamic DOM, computed styles, and JavaScript-heavy applications more realistically.

It is designed for teams that want a fast local check, stable JSON for automation, and a single binary that can be dropped into CI.

Why use it

• Real browser signals instead of static guesses
• Works for single pages, sitemaps, URL lists, and same-domain crawl discovery
• Outputs as terminal table, JSON, PDF, AI-optimized task list, or compact summary JSON for dashboards
• JSON output is schema-backed and tested for release stability
• Ships as a Rust binary instead of a Node-based toolchain

Why this approach

Most accessibility CLIs either depend on static parsing or require a heavier runtime stack around browser automation. auditmysite is opinionated in a different direction:

• Chrome-native accessibility data first
• CLI-first workflow for local use and CI
• Small operational surface: install a binary, point it at a URL, get a report
• Optional modules for performance, SEO, security, and mobile without changing tools

Quick Example

auditmysite https://example.com

By default, a single URL audit runs the full analysis set, prints a compact terminal summary, and writes report artifacts into the current working directory:

• ./example-com-YYYY-MM-DD-single-report.pdf
• ./example-com-YYYY-MM-DD-single-report.json
• ./example-com-YYYY-MM-DD-single-report-screen-reader-audit.json

For CI or machine-readable output:

auditmysite https://example.com -f json -o report.json --quiet

Install

curl installer (macOS/Linux)

curl -fsSL https://raw.githubusercontent.com/casoon/auditmysite/main/install.sh | bash

The installer downloads the latest GitHub Release asset for your platform and verifies it against the published .sha256 checksum before installing it.

Upgrading: run the same command again. The installer detects where your current binary lives and replaces it in place — no PATH conflicts, no leftover old version.

Note: If you previously installed via cargo install auditmysite, remove that binary first so the script installs to the right location:

rm ~/.cargo/bin/auditmysite
curl -fsSL https://raw.githubusercontent.com/casoon/auditmysite/main/install.sh | bash

Verify the installation:

auditmysite --version
auditmysite --help
auditmysite https://example.com

That default command writes report artifacts into the current directory, for example:

• ./example-com-YYYY-MM-DD-single-report.pdf
• ./example-com-YYYY-MM-DD-single-report.json
• ./example-com-YYYY-MM-DD-single-report-screen-reader-audit.json

cargo install (crates.io)

cargo install auditmysite

Builds and installs the binary from source. Tested with the pinned toolchain (see Toolchain); older compilers are not supported.

Prebuilt binaries

Download from Releases.

• macOS/Linux: .tar.gz
• Windows: .zip

Build from source

git clone https://github.com/casoon/auditmysite.git
cd auditmysite
cargo build --release
./target/release/auditmysite --version

Toolchain

rust-toolchain.toml pins the exact Rust version used for local development, CI, and release builds; rustup picks it up automatically inside the repository. There is no MSRV promise: the pin is raised deliberately, and older compilers are not supported.

Optional Cargo features:

Feature	What it adds	Build command
pdf	PDF report generation via the renderreport/Typst engine	cargo build --release --features pdf
pdf_test	PDF rendering integration tests	cargo test --features pdf_test
ai-transparency	C2PA image-provenance check (EU AI Act Art. 50, opt-in --ai-transparency flag, single-URL mode only)	cargo build --release --features ai-transparency

Requirements

• Rust as pinned in rust-toolchain.toml for local builds
• Chrome, Chromium, or a managed browser install (auditmysite browser install)
• macOS, Linux, or Windows for released binaries

auditmysite requires a browser to be present at run time. It does not download or install one automatically — it reports an error and exits if none is found. To install a managed Chrome for Testing into ~/.auditmysite/browsers/:

auditmysite browser detect                    # show what's found
auditmysite browser install --headless-shell  # smaller; fallback only (see below)
auditmysite browser install                   # download Chrome for Testing (opt-in)

On macOS, batches with keyboard journeys audit one page at a time when a full browser is used: Chrome runs the AppKit event loop even in headless mode, and keyboard events a page does not consume go through it on the browser's main thread; with several pages at once that can stall the whole browser. --concurrency overrides this. The headless shell is not affected, but sites with bot protection may serve it a challenge page, so it is only used when no system browser is found.

Quick Start

The fastest way to validate your setup:

auditmysite https://example.com

That creates the default report set in the current directory. For machine-readable output only:

auditmysite https://example.com -f json -o report.json

Single page

# default: full audit + terminal summary + PDF/JSON in current directory
auditmysite https://example.com

# JSON
auditmysite https://example.com -f json -o report.json

# PDF with explicit path
auditmysite https://example.com -f pdf -o report.pdf

# stricter WCAG level
auditmysite https://example.com -l AAA

Batch audits

# explicit sitemap
auditmysite --sitemap https://example.com/sitemap.xml

# crawl from a base URL and discover same-domain pages automatically
auditmysite https://example.com --crawl --crawl-depth 2

# base URL: probe robots.txt / common sitemap locations first
auditmysite https://example.com

# prefer sitemap automatically if one is found
auditmysite https://example.com --prefer-sitemap

# suppress sitemap suggestion and stay on the single page
auditmysite https://example.com --no-sitemap-suggest

# URL file
auditmysite --url-file urls.txt

# per-page reports: scan a list/sitemap but write one PDF per URL instead of an aggregated batch report
auditmysite --url-file urls.txt --per-page-reports --output reports/per-page/
auditmysite --sitemap https://example.com/sitemap.xml --per-page-reports --output reports/per-page/

Pages run in parallel (--concurrency). A page waiting for a free browser page waits as long as one page in flight may take (creating the page, four times --timeout with a two-minute minimum, and the page reset), so heavy pages ahead of it no longer push it out. A page that still gets no browser page in time is retried one at a time after the parallel phase. Pages that cannot be audited at all are not part of any score: the report says how many URLs the scores cover (JSON: summary.url_count of summary.attempted_url_count, failures under errors; PDF: cover and status section).

Browser selection

auditmysite --browser-path /path/to/chrome https://example.com

CLI

auditmysite [OPTIONS] [URL] [COMMAND]

Primary commands:

• auditmysite <url>: run a full single-page audit and write PDF/JSON into the current directory
• auditmysite --sitemap <url>: audit sitemap URLs
• auditmysite --url-file <file>: audit URLs from file
• auditmysite <url> --crawl: discover same-domain pages from a seed URL and audit them as a batch
• auditmysite browser detect: show available browsers
• auditmysite browser install [--headless-shell]: download and install Chrome for Testing or the headless shell into ~/.auditmysite/browsers/ (opt-in, never automatic; used only when no system browser is found)
• auditmysite doctor: run local diagnostics

Useful flags:

• --prefer-sitemap: if a sitemap is detected for a base URL, switch directly into batch mode
• --no-sitemap-suggest: suppress sitemap probing/suggestion and keep the run on the single URL
• --crawl-depth <n>: limit same-domain crawl discovery depth when using --crawl
• --per-page-reports: scan a URL list or sitemap but write one individual report per URL instead of an aggregated batch report; -o is treated as a target directory. With -f json, index.json and findings.jsonl are written next to the page files (see Technician mode)
• --technician: fix-oriented batch preset, see Technician mode
• --include-path <glob> / --exclude-path <glob>: select batch URLs by path before --max-pages (repeatable)
• --no-screen-reader-report: do not write the *-screen-reader-audit.json sidecar
• --html-conform: run the HTML-conformance module without --full
• --lang <de|en>: set the language for PDF reports (default: de)
• --stack: enable tech stack detection and stack-specific security probes (included automatically with --full)
• --interactive <off|basic|full>: control the Accessibility Journey Layer for interactive checks — tab walk, skip-link, modal focus trap, SPA navigation, form-error announcement, link-text inventory (default: full; use off for fastest runs)
• --annex <en301549|bik>: add an opt-in appendix to the PDF report. en301549 maps findings to EN 301 549 (chapter 9, "Web") clauses — a technical building block for a human-authored accessibility statement, not a statement itself. bik regroups the same findings by the chapters of the "BIK für Alle" editorial guide (images/alt text, link text, structure, easy language, PDFs, videos), without adding new checks. The underlying JSON data (en301549_annex, bik_guide) is always present regardless of this flag; it only gates the PDF section.
• --dns-check: opt-in, score-neutral DNS configuration check (CAA presence, best-effort DNSSEC, SPF when an MX record exists); runs once per host, not once per page
• --isolate-third-party-impact: reload the page once per top-5 third-party origin with that origin blocked and report each origin's Total Blocking Time impact; costly, requires --full or --performance, single-URL mode only
• --check-ssr-content: reload the page with JavaScript disabled and flag an SSR/hydration content gap when essential content only appears client-side; one extra reload, requires --full or --seo, single-URL mode only
• --display <calm|text|visual|all>: audit one display mode of the data-display convention — see Display modes
• --design-quality: opt-in UX/readability heuristics, including alt-text quality checks (filename-like alt text, "image of" prefixes, overly long or redundant alt text); score-neutral and not part of --full
• --exclude-selector <CSS> (repeatable): drop findings inside the subtree a CSS selector matches — for markup that is broken on purpose, such as teaching specimens (see Excluding intentional specimens)
• --color <auto|always|never> / --progress <auto|always|never>: terminal color and batch progress policy; --progress is independent of --quiet

For the full current interface, use:

auditmysite --help
auditmysite browser --help

Output Contract

JSON output is treated as an automation contract.

• Contract documentation: docs/OUTPUT_CONTRACT.md
• Single report schema: docs/json-report.schema.json
• Batch report schema: docs/json-batch-report.schema.json

Key fields in a single-page report:

• metric_context — machine-readable definitions for the 0–100 score scale and the report's scoped count fields
• findings — static WCAG violations and SEO findings
• interactive_findings — journey-phase results (link texts, landmarks, heading outline, focus order, modal traps …); present when --interactive basic|full was used
• accessibility_journey — structured trace of each journey (steps, snapshots, durations); present when --interactive basic|full was used
• audit_scope and execution_environment — requested modules, viewports, throttle profiles, interaction mode, browser context, and live/cache provenance
• audit_quality plus pages[].detail.module_runs and pages[].detail.rule_outcomes — distinguish complete, partial, failed, skipped, and non-applicable checks so a measurement failure cannot look like a clean result
• pages[].detail.accessibility_assessments — structured warnings, manual-review items, and positive signals kept separate from confirmed violations and scoring
• artifacts — descriptors for separately written evidence or screen-reader sidecars without embedding binary data in the main JSON
• build_id — git short-SHA of the binary that produced the report (suffixed -dirty for builds from uncommitted changes, unknown for builds without a .git directory), so two reports with the same tool_version can be told apart
• pages[].detail.en301549_annex and pages[].detail.bik_guide — findings mapped to EN 301 549 clauses and to the "BIK für Alle" editorial guide chapters; always present, independent of --annex
• pages[].detail.accessibility_subcategory_scores and pages[].detail.security_category_scores — the Accessibility and Security scores broken down by subcategory ({ name, score }), the same breakdown the PDF shows; single reports only

Rule IDs are stable across releases. Version 1.3.0 renamed two IDs that collided with unrelated checks: the 4.1.2 control-label check is now control-missing-label (was label, which stays with the 3.3.2 label/instructions checks), and the 2.4.9 link-purpose (link only) check is now link-name-only (was link-name, which stays with the 2.4.4 check).

For dual-viewport audits, the Accessibility score is the rounded blend of 70% mobile and 30% desktop in both JSON and PDF. WCAG occurrences, distinct grouped WCAG findings, and findings from all categories are exposed separately so counts remain comparable across formats.

Batch JSON additionally exposes site_analysis: module averages, consistency signals, page types, topic overlap, duplicate content, structured-data distribution, performance rollups, interactive coverage, and aggregated accessibility assessments. Per-page entries remain compact instead of duplicating full single-page reports.

The repository validates these contracts in automated tests.

Feature Scope

WCAG rules (Level A and AA)

Core rules:

• Non-text content (1.1.1)
• Keyboard access (2.1.1)
• Bypass blocks (2.4.1)
• Language of page (3.1.1)
• Name, role, value / form labeling (4.1.2)
• Contrast minimum (1.4.3) and non-text contrast (1.4.11)
• Headings and labels (2.4.6)
• Labels or instructions (3.3.2)
• Focus order (2.4.3), focus visible (2.4.7), and focus not obscured, minimum/enhanced (2.4.11/2.4.12, WCAG 2.2)
• Label in name (2.5.3)
• Meaningful sequence — CSS order vs. reading-order mismatches (1.3.2)
• Pause, stop, hide — <marquee> and long-running CSS animations without a pause control (2.2.2)
• Redundant entry — same field requested twice with no reuse/autofill hint (3.3.7, WCAG 2.2)
• Accessible authentication — paste blocked on password or one-time-code fields, measured with a synthetic paste event (3.3.8, WCAG 2.2); CAPTCHAs in sign-in forms as review items
• Target size minimum (2.5.8, WCAG 2.2) and text spacing (1.4.12)

ARIA and semantics:

• ARIA role validation — invalid roles, required owned elements, required context
• ARIA attribute checks — allowed attributes per role, required attributes, prohibited attributes
• Accessible name checks — icon-only controls, empty aria-labelledby/describedby, name/description conflicts, naming by role type (command, input, meter, progressbar, toggle, dialog, treeitem)
• ARIA relationship checks — aria-controls, aria-owns, aria-activedescendant, duplicate IDs
• Landmark structure — main, navigation, banner, contentinfo (presence, uniqueness, top-level nesting, no-duplicate for banner/contentinfo/main, required parent for landmarks)
• Content in landmarks — region rule ensuring body content lives inside landmark regions
• Table rules — caption/name, header cells, presentational tables, cell placement
• Form rules — fieldset/legend for grouped controls, required field indication, error description, label-title-only detection
• List structure — listitem context, empty lists, definition list integrity
• Dialog rules — accessible name, aria-modal, alert region labeling
• Widget rules — tab/tabpanel pairing, selected state, combobox options, slider value, tree context, summary element naming
• Media rules — application and image-role elements without accessible names
• Video checks — caption tracks for native <video> (a same-origin track file is probed before a pass is confirmed, 1.2.2), media alternatives with a nearby-transcript heuristic (1.2.8), and keyboard operability of native video controls (unnamed or unreachable players)
• ARIA hygiene (best practice, low severity) — explicit roles that restate the element's implicit role, and links without a real target (href="#", javascript:) used as button substitutes
• Duplicated accessible names — names that repeat themselves back to back (e.g. "Contact Contact"), typically from an icon label concatenated with adjacent text
• Frame and iframe rules — accessible names on all frames (frame-title), manual-review notices for cross-origin frames (frame-tested), and a full WCAG content scan inside same-origin iframes: image-alt (1.1.1), button-name (4.1.2), link-name (2.4.4), form labels (1.3.1), referenced duplicate IDs (4.1.2), document language attribute (3.1.1)
• SVG rules — SVG image accessible names
• Server-side image maps — detection and flagging
• Meta viewport — large maximum-scale restrictions

100+ rules with stable rule_id, tags (e.g. wcag2a, wcag412, cat.aria), and an impact field (critical / serious / moderate / minor).

Methodology numbers are frozen in docs/PARITY_CONTRACT.jsonc and guarded by tests/parity_contract.rs: WCAG 2.2 AA has 55 A/AA criteria, 30 are covered by automated AuditMySite checks (a criterion counts only if its rule can report a violation), and 22 are listed as manual-review criteria.

Some criteria (keyboard trap behavior, timed content, captions) cannot be reliably verified by automated means. These are flagged as not_testable in the JSON output and listed in the report's audit scope section as requiring manual review.

AAA is not fully implemented yet.

Additional modules

Modules are classified by how their result is obtained: compliance (Accessibility), measured (based on real browser data), composite (the search-experience roll-up, which blends measured SEO with heuristic sub-scores), heuristic (structural-signal estimates, marked with ~ in reports), and optional (Dark Mode — reported as a design choice, not a compliance gap).

Measured:

• Performance: Core Web Vitals (FCP, LCP, TBT, CLS), throttled profiles, DOM/load targets, render-blocking and third-party resources, critical request chains, unused code, minification potential, JavaScript heap, and modeled transfer emissions
• SEO: meta tags, headings, structured data, page-to-schema fit, content profile, tracking/external services signals, and social-preview images (og:image/twitter:image) without an alt description
• Security: HTTPS, header checks, and CDN/WAF protection detection. Headers are grouped by risk tier: CSP, HSTS, and clickjacking protection are baseline requirements, while context-dependent headers such as COOP/CORP get a verification question and safe-configuration guidance instead of an unconditional "add this header"
• Mobile: viewport, touch-target, readability checks, UX heuristics (cookie-banner, modal/overlay, CTA detection)
• HTML5 conformance: spec-conformance checking via the html-conform crate, part of --full; scored per distinct defect cause rather than per raw occurrence, so one templated markup mistake rendered many times doesn't get charged once per render
• DNS configuration (opt-in --dns-check): CAA, best-effort DNSSEC, and SPF when an MX record exists; score-neutral

Heuristic (indicator scores — tendency, not measurements):

• UX: 5-dimension analysis (CTA clarity, visual hierarchy, content clarity, trust signals, cognitive load) with saturation curve scoring
• Journey: user-flow analysis (entry clarity, orientation, navigation, interaction, conversion) with page-intent-aware weighting
• AI Visibility: structural readiness for LLM indexing and citation (readability, citability, structured data, AI policy, chunk quality)
• Source Quality: code hygiene signals (inline styles, deprecated elements, semantic structure, asset hygiene)
• Dark Mode: detects dark mode support via prefers-color-scheme media queries and CSS custom properties
• Easy language: recognizes an easy-language ("Leichte Sprache") or plain-language version of the page via class names, a lang variant, or a link-text marker, and reports it as a positive signal
• Tech Stack: detects CMS and frameworks (WordPress, Drupal, Joomla, Next.js, Astro, React, Vue, etc.) via in-page signals and runs stack-specific security probes (admin panel exposure, user enumeration, version disclosure)
• Commerce: shop audit that only activates when a page is detected as a store (schema-gated). Checks product structured-data completeness, presence of mandatory and trust pages (imprint, returns, shipping, payment), coarse page-kind classification (product detail, category), and rolls findings up across a batch. Derive-only — no extra browser interaction. Product-detail pages also get two commerce-aware interactive journeys — see Accessibility Journey Layer below.

Display modes

Sites can offer display modes so that nobody depends on 3D, animation or visualisations. The BarrierLab data-display convention (draft v0) sets <html data-display="visual|calm|text"> before the first paint, stores the visitor's choice in localStorage under the key display, marks each visualisation as figure[data-viz="chart|diagram|3d|image|interactive"] with a text layer [data-viz-text], and puts a [data-display-toggle] control on every page with a visualisation.

• Detection (always on): every page that sets html[data-display] or contains a figure[data-viz] reports pages[].display_modes in the JSON — offered modes, the mode it rendered in, whether the attribute was set before <body>, the toggle, and the visualisations per kind. The PDF shows it as "Page display modes: visual · calm · text".
• --display calm|text|visual: before navigation the choice is stored in localStorage.display (Page.addScriptToEvaluateOnNewDocument), and calm/text also emulate prefers-reduced-motion: reduce (Emulation.setEmulatedMedia) — so sites without the convention that honour the media query get their reduced variant too. Without the flag the site default is audited, as before. audit_scope.display_mode in the JSON (site_default, visual, calm, text) and the PDF name the mode the scores belong to.
• --display all: audits each mode as its own run and writes one report per mode (report-calm.pdf, report-text.pdf, report-visual.pdf; batches: one batch report per mode). There is never a blended score. visual runs only for pages with figure[data-viz="3d|interactive"], with twice the page timeout. Findings are not cross-marked as "occurs in all modes" — compare the per-mode reports.
• Convention checks display/* and viz/* (best-practice, not WCAG requirements; anchored to 2.2.2, 1.1.1 or 1.3.1). From the markup, shared with a11y-rules: display/toggle-missing (visualisations without a toggle), viz/text-missing (no or an empty [data-viz-text]), viz/caption-missing (no <figcaption>), viz/static-missing (data-viz="3d"/"interactive" without [data-viz-static]), viz/table-missing (a chart without a <table>; a review note). Measured on the running page: display/init-missing (data-display missing or set only after <body> started, measured by an observer injected before navigation), display/text-media-visible (in text mode a visualisation still shows canvas/SVG/video/static picture), display/text-not-visible (in text mode a visualisation's [data-viz-text] has content but is not visible), display/text-hidden ([data-viz-text] removed from assistive technology by hidden, aria-hidden, inert or CSS, in any mode). The text-mode checks key on the mode the page actually rendered in.

auditmysite --sitemap https://example.com/sitemap.xml --display calm --format pdf --output reports/example-calm.pdf
auditmysite https://example.com/page --display all --output reports/example-page.pdf

Structured-data analysis

The SEO module parses JSON-LD objects, arrays, and @graph documents; normalizes short, multiple, and full-IRI @type values; and reports invalid JSON, missing or invalid context, and untyped nodes. Microdata and RDFa are detected and explicitly marked as detected but not content-validated.

Type-specific rules assess Product/Product Snippet, merchant Product + Offer, Article/BlogPosting/NewsArticle, BreadcrumbList, Organization, LocalBusiness, FAQPage, Event, Recipe, VideoObject, JobPosting, SoftwareApplication/WebApplication/MobileApplication, ProfilePage, CollectionPage/ItemList, WebPage/WebSite, and Person. Each profile records its source and review date. Eligibility blockers, recommendations, and manual checks remain separate; unknown types stay visible in the inventory without being judged incomplete.

Page-to-schema fit is evaluated conservatively from visible page intent, visible facts, and URL evidence. The tool distinguishes product, service/software, job, event, FAQ, person, location, editorial-review, corporate, hub, and lead-generation pages. Missing primary schema is only reported as an opportunity at high classification confidence, and single-item Product, JobPosting, or Event markup is rejected on corresponding overview routes.

For supported types, visible titles, prices, availability, authors, dates, FAQ content, breadcrumbs, job titles, and event dates are compared with JSON-LD. A hard mismatch is emitted only when the visible value is unambiguous; otherwise the result explicitly remains not evaluated or requires manual review. Batch reports additionally show recurring schema blockers, page-type/schema combinations, content-parity mismatches, and conflicting Organization/WebSite identities.

Runtime and evidence reliability

Page capture uses a bounded stability budget and records whether the DOM became quiet, an application-provided ready signal was observed, or the budget expired. Consent handling reports detected, dismissed, failed, and unknown states with non-sensitive evidence. Ctrl-C and SIGTERM follow the same controlled shutdown path as normal runs, and report files are written atomically so partial files are not presented as successful output.

Every audited page receives Do-Not-Track and Global Privacy Control signals (DNT: 1 and Sec-GPC: 1 on all requests, plus the matching navigator.doNotTrack/navigator.globalPrivacyControl values), so an audit visit is less likely to show up in the site owner's analytics. This is best-effort: not every analytics tool honors these signals.

Sitemap indexes are deduplicated and guarded against cycles, with hard limits of 1,000 sitemap documents and 100,000 discovered URLs. Batch aggregation stays bounded and publishes its atomic report only after collection succeeds.

Accessibility Journey Layer

Interactive checks run a real browser session after the static AXTree phase. They run in full mode by default and can be reduced via --interactive <off|basic|full> or mode in auditmysite.toml.

Mode	What runs
off	No interactive phase — fastest, no browser interaction after initial load
basic	Tab-walk (focus order, reverse jumps), skip-link verification, disclosure/accordion, modal focus trap, tab-list, menu journey
full (default)	Everything in basic, plus: SPA-navigation detection, form-error announcement (now covering multiple independent forms per page, e.g. search + login + newsletter, and flagging live regions inserted only after submit, which not every browser/screen-reader combination announces), link-text inventory (generic/duplicate texts, heading outline, landmark structure)

On a detected shop's product-detail page, full mode also runs two commerce-aware journeys: an add-to-cart feedback check (does adding an item announce the result via a live region or focus-managed dialog, or only update a visual cart badge — SC 4.1.3) and a quantity-stepper operability check (can the quantity field be operated by keyboard, and does its value stay exposed to assistive technology — SC 2.1.1/4.1.2). Both are click-only, single-interaction checks — never a real checkout submission, never a filled-in purchase form.

Results appear in interactive_findings and accessibility_journey in the JSON output. The execution block records detected, attempted, completed, failed, skipped, and budget-limited journeys separately from findings. Compact focus evidence retains visibility, viewport, focus-indicator, bounding-box, obscuring, aria-hidden, and inert signals without embedding a full AXTree. Interactive findings do not affect the accessibility score or legal_flags; critical interactive findings can raise the risk level.

auditmysite.toml configuration:

[interactive]
mode = "full"             # off | basic | full
journey_budget_ms = 8000  # wall-clock budget per URL in milliseconds (default: 6000)

Risk assessment

Risk level is computed independently from the score. A page scoring 81 can still carry "Critical" risk if it has Level A violations relevant under BFSG/EAA. Risk levels: Low, Medium, High, Critical — based on critical/high violations, legal flags, and blocking issues (4.1.2/2.1.1).

Configuration file

auditmysite.toml is an optional project-level config file placed in the working directory. It supports [audit], [rules], [interactive], [thresholds], and [budget] sections.

Rule configuration

Rules can be selectively disabled or filtered via auditmysite.toml:

[rules]
disabled = ["heading-order", "landmark-one-main"]
# enabled_only = ["image-alt", "label"]  # run only these rules

Excluding intentional specimens

Sites that teach accessibility ship examples that are broken on purpose. Like axe-core's exclude, auditmysite can leave such regions out of the findings — never a page, never a rule:

• data-audit-exclude is always honoured. Put it on an element and its whole subtree is excluded: <section data-audit-exclude>…specimen…</section>.

• --exclude-selector <CSS> (repeatable) excludes the subtree of every element the selector matches, e.g. --exclude-selector '[data-specimen]'. The same list can live in auditmysite.toml:

  [audit]
  exclude_selectors = ["[data-specimen]"]

The page is still audited in full; only findings whose element lies inside an excluded subtree are dropped before scoring (WCAG violations and warnings, pattern findings, and journey findings that name their element). Page-level findings are never excluded, and a finding located only by a selector is dropped only when every element that selector matches lies inside an excluded subtree.

Excluding never happens silently. The JSON report lists, per page, every applied selector with the number of elements it matched — including 0 and invalid selectors — and how many finding occurrences were dropped, per rule (pages[].exclusions; batch totals in summary.exclusions). The PDF names the selectors and counts in the methodology section (batch: in the audit frame on the cover).

AI / LLM output format

Export findings as a task-oriented JSON list for direct LLM processing:

auditmysite https://example.com -f ai -o findings.json

Each entry is a task object with task_id, rule_id, impact, wcag, tags, title, issue, fix, selector, node_id, and help_url — sorted by impact severity. Suitable for direct use as context in AI-assisted code remediation.

Baseline and CI diff

Save a baseline snapshot and compare future runs against it:

# Save baseline
auditmysite https://example.com -f json -o baseline.json

# Future CI runs can diff against the baseline programmatically via the Rust API

The Baseline type in the audit module supports from_violations, diff, load, and save.

Non-goals (deliberately deferred)

These are recognized as potentially useful but deliberately not on the roadmap right now — each would pull the project away from its core shape (a stateless, single-run web-page auditor) for a disproportionate amount of new surface area:

• PDF document accessibility checking. auditmysite audits rendered web pages via the browser's accessibility tree; it does not parse or score linked/embedded PDF documents (tagged-PDF structure, /Lang, alt text on figures, table header scope, etc.). Doing this properly would need a new PDF-parsing dependency and a second, PDF-specific SSRF-hardened fetch path (comparable in scope to the existing ai_transparency module) for a feature that only applies when a page happens to link PDFs. Out of scope for now; revisit only as a dedicated, separately-scoped module if real demand shows up.
• Full audit-history / report-diffing as a first-class feature. The Baseline API above covers programmatic before/after diffing of WCAG violations for a single URL. A richer, built-in "compare this report against an older report" mode with its own CLI flag, PDF section, and batch-vs-single/methodology-consistency handling is not planned — that turns the tool from a stateless auditor into a report-history manager, a different product shape. A JSON-to-JSON diff against two saved reports is straightforward to script externally (e.g. with jq) without needing this built in.

Report Modes

Single-page reports and sitemap/batch reports are intentionally different.

Single-page report is a product-grade PDF organized as a top-down narrative:

• Cover: a composed dashboard — dominant overall score with a score-band label (no A–F grade, no "/100"), a module gauge strip, and the WCAG findings scope.
• Management view: severity counters, a "quality profile" spider radar, a score-driver table (which weighted module pulls the overall score down most), Accessibility and Security subcategory breakdowns, a one-line problem profile (isolated, concentrated in a few systemic patterns, or broad and systemic), and strengths / optimization cards.
• Findings overview: a compact, priority-sorted findings matrix precedes the detailed finding cards.
• Manual review: a fixed list of WCAG criteria that are structurally outside automated testing, and concrete VoiceOver/NVDA self-test steps where a quick manual check is realistic.
• Accessible output: PDFs are tagged and PDF/UA-1 conformant (structure tree, document language, title, bookmarks); rendering fails with a diagnostic instead of silently producing a non-conformant PDF.
• Module chapters: each module is its own chapter with a magazine-style opener and a one-line key takeaway. AI Visibility, Content Visibility, and Source Quality are merged into a single "KI & Vertrauen" (AI & Trust) chapter.
• Action plan: recommendations as action cards grouped by where the problem lives (systemic vs. local), without time or effort estimates, plus a root-cause distribution chart.
• Evidence-grade findings: each finding card can include a cropped, highlighted screenshot of the affected element, its DOM path, and (where applicable, e.g. contrast) the measured vs. required value — so a finding stands on its own without re-running the tool.
• Audit coverage: requested and completed checks, partial measurements, manual-review items, and Journey execution coverage are surfaced explicitly instead of treating missing data as a pass.
• Performance decisions: raw resource and loading metrics are paired with target ranges, the largest directly actionable lever, and prioritized actions.

The design follows a consistent four-color status system; reports use no emoji and report effort by priority rather than by time windows.

Sitemap/batch report is aggregated and domain-wide: averages, ranking, recurring issues, URL matrix, near-duplicate content, broken links, crawl diagnostics. It also verifies which recurring findings share the same underlying template component across pages — reporting "one fix resolves N pages" instead of N near-identical findings, with a confirmed/likely confidence distinction so the claim is never overstated. Template clusters require at least three affected pages and 60% site coverage, can identify selector-less document findings, and retain header/nav/main/footer context. Cross-page WCAG assessments for consistent navigation, identification, help, and multiple ways are kept separate from single-page automation and explicitly mark evidence gaps as manual review.

Batch reports are not a stack of single-page reports.

Compared to typical setups

• Better fit for JavaScript-heavy sites than static HTML-only checks
• Easier to distribute than a multi-package browser toolchain
• More automation-friendly than ad hoc console output because the JSON contract is explicit and tested
• Broader reporting surface than a pure accessibility-only checker when you also want performance, SEO, security, and mobile signals
• Violations carry stable rule_id, tags, and impact — easier to integrate with existing tooling or dashboards

Screen Reader Audit vs. axe-core / Pa11y

Standard accessibility checkers verify individual rules in isolation. auditmysite additionally simulates the sequential experience of a screen reader user navigating the page — detecting problems that only emerge in context.

Capability	axe-core	Pa11y	auditmysite
Rule-based WCAG checks	✓	✓	✓
Reading sequence simulation	—	—	✓
Out-of-context link text analysis (duplicate "Read more" × 8)	—	—	✓
Accessible name quality score (not just present/absent)	—	—	✓
Landmark navigation strategy (can a SR user reach main content?)	—	—	✓
BFSG / EN 301 549 legal mapping per finding	—	—	✓

Every finding also carries a per-criterion EN 301 549 (chapter 9, "Web") clause reference. A structured version — all 50 WCAG 2.1 A/AA clauses split into "violations found", "no violations in the automated scope", or "manual review required", plus which chapters (5–8, 10–13) sit outside this tool's audit scope entirely — is always in the JSON (en301549_annex) and can be added to the PDF as an appendix with --annex en301549. This is explicitly not an accessibility statement and doesn't claim to be one — it's a technical building block for a human-authored one, with an explicit scope disclaimer in both languages.

When the screen reader module runs, a JSON sidecar is written automatically next to the primary report:

example-com-YYYY-MM-DD-single-report.pdf
example-com-YYYY-MM-DD-single-report-screen-reader-audit.json  ← automatic sidecar

The sidecar models what a screen reader would typically announce, node by node, based on rule-based conventions (name, role, and state) — not a verified 1:1 reproduction of what NVDA, JAWS, or VoiceOver actually announce, which can vary by assistive technology, browser, and locale. It also flags which announcements are ambiguous or missing, suitable as a developer reference and as supporting evidence for BFSG compliance audits. No extra flag is required; the file is created whenever screen reader data is available in the audit result.

Typical Workflows

Examples grouped by audience and goal.

Customer-facing report (PDF)

Single-URL audit with full module coverage and a custom logo on the cover.

# default: writes a PDF + JSON sidecar to the current directory
auditmysite https://example.com --full

# explicit branding and output path
auditmysite https://example.com --full --logo ./assets/customer-logo.svg --output reports/customer.pdf

# pick a report depth: executive (management), standard (default), technical (developers)
auditmysite https://example.com --full --report-level executive --output reports/exec.pdf

# PDF language (default: de)
auditmysite https://example.com --full --lang en --output reports/report-en.pdf

CI / automation (JSON)

Quiet, machine-readable output for pipelines.

# exit code follows score thresholds; JSON report for downstream tooling
auditmysite https://example.com -f json -o report.json --quiet

# batch CI run on a sitemap
auditmysite --sitemap https://example.com/sitemap.xml -f json -o sitemap-report.json --quiet

AI fix list

Compact, agent-friendly output that focuses on actionable fixes.

auditmysite https://example.com -f ai -o fixes.json

Dashboard / ranking feed

Compact summary JSON with score, grade, medal, issue counts, and top 10 findings — matches the lastAudit schema used by dashboard tools.

auditmysite https://example.com -f summary -o summary.json

Sitemap / batch

Domain-wide audits with cross-page aggregation.

# explicit sitemap
auditmysite --sitemap https://example.com/sitemap.xml --full

# crawl from a base URL
auditmysite https://example.com --crawl --crawl-depth 2 --max-pages 50 --full

# URL list from file
auditmysite --url-file urls.txt --full

# one PDF per URL instead of an aggregated batch report
auditmysite --sitemap https://example.com/sitemap.xml --per-page-reports --output reports/per-page/

Technician mode

For people who fix the issues rather than read a report: one JSON file per page plus two flat files to script against, no PDF.

auditmysite --sitemap https://example.com/sitemap.xml --technician -o reports/tech/
auditmysite --url-file urls.txt --technician -o reports/tech/

# only part of the site, before -m applies
auditmysite --sitemap https://example.com/sitemap.xml --technician \
  --include-path '/blog/**' --exclude-path '/blog/tag/**' -m 50 -o reports/tech/

--technician is shorthand for --per-page-reports -f json --no-screen-reader-report --seo --html-conform. It runs the modules that produce fixable findings — accessibility (including the keyboard journeys; --interactive off makes it faster), HTML conformance and SEO — and skips the throttled performance passes, mobile, security and tech-stack detection. Each page file records that partial scope in execution.scope.requested_modules and execution.module_runs. Flags you give explicitly win: -f replaces the format, and --full, --performance, --mobile, --security add their modules as usual.

The output directory then holds:

• one <site>-<path>-<date>-single-report.json per audited page (the regular single-page JSON),
• index.json: every attempted URL in input order with file (or null), status (ok, blocked for bot walls/access denials, failed), reason, overall_score, accessibility_score, finding_count, occurrence_count and audit_quality (schema: docs/technician-index.schema.json),
• findings.jsonl: one line per finding occurrence across all pages — url, source (wcag, journey, seo, html_conform), rule_id, wcag_criterion, level, severity, selector, location, message, fix_suggestion, viewport_tags (schema: docs/technician-finding.schema.json). Unlike the page files, which keep a few example occurrences per finding, this list is complete.

Any --per-page-reports -f json run writes index.json and findings.jsonl; failed or blocked pages appear only in index.json, never as a page file. All text is canonical English.

Path globs are matched against the whole, percent-decoded URL path (no host, no query): * and ? stay within one path segment, ** crosses segments, and /**/ also matches a single /. /blog/** selects everything below /blog/ but not /blog itself. A URL is audited when it matches any --include-path (or none is given) and no --exclude-path. With --crawl, the filters apply to the discovered pages; discovery itself is still capped by -m.

cd reports/tech
# the worst rules across the site
jq -r '"\(.severity)\t\(.rule_id)"' findings.jsonl | sort | uniq -c | sort -rn | head
# every critical/high occurrence with page and selector
jq -r 'select(.severity=="critical" or .severity=="high") | [.url, .rule_id, .selector // .location] | @tsv' findings.jsonl
# all occurrences of one WCAG criterion
jq -c 'select(.wcag_criterion=="1.4.3") | {url, selector, message}' findings.jsonl
# pages that were not audited, and why
jq -r '.pages[] | select(.status!="ok") | "\(.status)\t\(.url)\t\(.reason)"' index.json
# pages by accessibility score, lowest first
jq -r '.pages[] | select(.status=="ok") | "\(.accessibility_score)\t\(.occurrence_count)\t\(.url)"' index.json | sort -n

Local development

# audit a local dev server with a system Chrome
auditmysite https://localhost:3000 --browser-path /Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome

# quick CLI summary without writing files
auditmysite https://example.com --format table

Base URL with sitemap suggestion

# interactive: ask first if a sitemap is found
auditmysite https://example.com

# non-interactive: switch directly to sitemap mode
auditmysite https://example.com --prefer-sitemap

# stay on the single URL even when a sitemap exists
auditmysite https://example.com --no-sitemap-suggest

Architecture

CLI -> Browser Manager -> Chrome/CDP -> Accessibility Tree -> WCAG Engine -> Output

Key layers:

• browser/: browser detection, resolution, explicit install (browser install command only — no auto-download), lifecycle, pooling
• audit/: pipeline, normalization, scoring, batch processing
• wcag/: rule engine and violations
• output/: CLI, JSON, PDF, AI, summary format
• seo/, security/, performance/, mobile/, ux/, journey/: optional analysis modules
• tech_stack/, source_quality/, ai_visibility/, dark_mode/: heuristic indicator modules

Shared libraries from barrierlab

Parts of what auditmysite checks no longer live in this repository. They moved to barrierlab, a Rust monorepo of accessibility and web-conformance libraries, so that the same checks, rule IDs and wording serve auditmysite, astro-post-audit and liveaudit alike. auditmysite pulls them in as published crates.io versions:

Crate	Provides
a11y-rules	the shared WCAG rules (lists, headings, landmarks, IDs, language, …)
a11y-dom	the document model those rules run on
a11y-report	the finding, outcome and rule-run model
accname	accessible name computation (accname 1.2, HTML-AAM)
a11y-perception	accessibility tree, snapshot, reading-order projection and snapshot diff
web-checks	robots.txt rules and bot classification, meta lengths, OpenGraph and JSON-LD structured-data checks
html-conform	HTML5 conformance checking, comparable to the W3C validator

This is a real dependency: a fix or a new rule in one of these areas is made in barrierlab, released there, and then taken over here by raising the version. auditmysite does not patch or fork them. What stays here is everything specific to this tool: its own WCAG rules, the keyboard journeys, the Chrome/CDP capture, scoring and taxonomy, the PDF report, the CLI and the report texts. For work on both sides at once, point [patch.crates-io] in a local, uncommitted .cargo/config.toml at a barrierlab checkout; commits always build against the published versions.

More detail:

• Current implementation: docs/ARCHITECTURE.md
• Browser dependency details: docs/chrome-dependency.md
• Troubleshooting: docs/TROUBLESHOOTING.md

Development

Setup

git clone https://github.com/casoon/auditmysite.git
cd auditmysite
cargo test
cargo build --release
./target/release/auditmysite https://example.com

Pre-commit checks

This repository uses Git hooks with a fast local pre-commit gate and a full pre-push gate.

pre-commit runs:

• nosecrets on staged changes
• cargo fmt -- --check
• cargo clippy --lib --bins --all-features -- -D warnings

pre-push runs:

• scripts/check-version-match.sh for pushed v* tags
• cargo clippy --all-targets --all-features -- -D warnings
• cargo test

Enable the repo hook path:

git config core.hooksPath .githooks

Install nosecrets as a real binary first:

npm install -g @casoon/nosecrets
# or
cargo install nosecrets-cli

Skip the Rust checks only when you intentionally need to bypass them:

SKIP_RUST_CHECKS=1 git commit -m "..."

The hook expects nosecrets to be available in PATH.

Debugging report content (hidden --debug-typ)

PDF reports are rendered through the renderreport/Typst engine. To review report completeness and wording without opening the binary PDF, use the hidden --debug-typ flag together with --format pdf. It writes the intermediate Typst source as a .typ sidecar next to the PDF, for both single and batch reports:

# Single report → reports/example-audit.pdf + reports/example-audit.typ
./target/release/auditmysite https://example.com --full --format pdf \
  --output reports/example-audit.pdf --debug-typ

# Batch report → reports/example-batch.pdf + reports/example-batch.typ
./target/release/auditmysite --sitemap https://example.com/sitemap.xml --full \
  --format pdf --output reports/example-batch.pdf --debug-typ

The .typ file is plain text and diff-friendly — useful for checking which audits land in the report and reviewing the exact wording of every section. The flag is intentionally hidden from --help (developer/debug use only).

Release checks

Run the local release gate with:

./scripts/release-check.sh

It validates:

• cargo test
• ignored browser integration tests
• builds with and without PDF
• current --help output
• JSON contract tests
• installer/release artifact consistency
• stale docs references

Troubleshooting

• Browser not found: run auditmysite browser detect or install a managed browser with auditmysite browser install
• "Access to '…' was blocked": the site answered HTTP 401/403/407/429 or served a bot challenge (Cloudflare, Fastly, Akamai, DataDome, Imperva, HUMAN). That page is not the site's content, so the URL is not scored; in a batch it is listed under errors. Ask the site owner to allow the auditing machine
• Running in Docker or as root: use --no-sandbox
• Need raw output for scripts: prefer -f json -o report.json
• Unsure about the full CLI surface: run auditmysite --help

Contributing

Library / Development

For library development or local work from the repository:

cargo build
cargo test

If you want the current local repository state as an installed binary while developing:

cargo install --path . --force

Contributions are welcome. At minimum before opening a PR:

cargo test
./scripts/release-check.sh

License

auditmysite is licensed under the MIT License. Use it freely, including commercially.

The shared rule engine, document model, accessible name computation, report model and HTML conformance checker live in barrierlab (see Shared libraries from barrierlab) and are shared with astro-post-audit and liveaudit — the same rule IDs across build time, CI and the live page.

MIT applies from version 1.5.0 onward. Earlier releases remain under the license that applied at the time: up to 0.25.x AGPL-3.0-or-later, 0.26.0 through 1.4.0 Business Source License 1.1. See NOTICE.

Credits

• Browser automation via chromiumoxide
• PDF reports via renderreport
• WCAG reference material from W3C