Changelog
Release history for crawwwl. Dates are when the version was published.
v0.2.114 May 2026
- CLI: new interactive
crawwwl assistantcommand — chat with AI about your audit findings directly in the terminal - CLI: new after an audit, a hint suggests running crawwwl assistant to discuss findings
- Dashboard: new audit assistant side panel with streaming AI chat, full audit context, and dynamic suggestions
- Dashboard: smooth accordion open/close on issue rows using CSS grid animation
- Dashboard: dimmed surrounding issues when one is expanded for readability
- Dashboard: issue accordions now show "View details" / "Close" instead of chevrons
- Dashboard: duplicate occurrence count removed from issue titles
- Dashboard: chat panel overlays without shifting audit content, margins preserved
- Dashboard: faster page navigation with loading states and view transitions
- Dashboard: feedback button styling simplified
- API: new streaming chat endpoint for CLI assistant
v0.2.06 May 2026
- CLI: new select depth and effort prompt redesigned with numbered choices, purple highlight, and full descriptions of what each audit level does
- CLI: new section separators between every output section for improved readability
- CLI: new long lines now wrap with hanging indent, aligned under the start of the text value
- CLI: update notification shown in a bordered box with link to changelog
- CLI: quick audit file cap removed — all scales now scan all project files
- CLI: crawwwl usage now shows balance only
- CLI: add usage link is now clickable in supported terminals
- API: fixed /v1/me 500 error caused by removed credit_balance column
v0.1.95 May 2026
- CLI: audit output now includes a disclaimer that crawwwl uses AI and findings should be verified before acting on them
- CLI: "High priority" section header now reads "Critical fixes" only when critical-severity issues are present
- CLI: suppress hint shown once per section instead of once per issue
- CLI: quick win always targets the highest-severity finding rather than the easiest fix
- Scanners: CSS module scanner now tracks per-alias imports, eliminating phantom class false positives when a component imports multiple CSS modules
- Scanners: SEO scanner no longer runs static metadata checks on pages that export generateMetadata
- crawwwl AI: architectural notes now require findings across at least three distinct files — single-file patterns no longer qualify
- crawwwl AI: false positive filtering improved with rule-specific dismissal criteria and stricter evidence requirements
v0.1.815 April 2026
- CLI: crawwwl usage command showing current balance, usage percentage, and a link to top up
- CLI: issue count summary moved directly below the score bars and reformatted to HIGH (1) / MEDIUM (0) / LOW (0)
- CLI: AI token usage line moved to the bottom of audit output with a link to crawwwl usage
- CLI: "(AI analysis unavailable)" now appears only when AI did not run
- API: /v1/me endpoint returning balance and usage percentage for authenticated keys
- Docs: CLI reference updated to v0.1.7 with the crawwwl usage command
v0.1.715 April 2026
- CLI: version is now injected at build time, so the header always shows the correct release
- CLI: update notification when a newer version is available on npm
- CLI: CTRL+C now cancels the running audit, so cancelled runs are not billed
v0.1.615 April 2026
- Scanners: all five scanners now use AST-based analysis, reducing false positives across performance, security, code quality, SEO, and accessibility
- Security: variable tracking improves detection of SQL injection, command injection, and XSS
- Accessibility: better checks for label associations, aria-expanded state, and keyboard support on button-like elements
- Performance: better handling for Next.js static image imports, raster vs SVG detection, and blocking scripts
- SEO: improved Next.js App Router metadata analysis, redirect page detection, and nested layout handling
- crawwwl AI: duplicate insight title/detail rendering and empty insight sections removed
- crawwwl AI: deep audit now retries malformed AI responses automatically, so insights still come through
v0.1.513 April 2026
- Usage: token usage is now logged for each audit, including crawwwl AI output tokens
- CLI: the default audit mode is now deep, with crawwwl AI included automatically
- Billing: Stripe top-ups in fixed 5, 10, or 20 amounts
v0.1.412 April 2026
- CLI: sign-in prompt during audit with options to sign in, continue without an account, or create one
- CLI: device auth no longer shows a confirmation code, and browser confirmation is now one step
- Dashboard: audit list score colours, a Latest badge, and source filters for all, GitHub, and CLI audits
- Dashboard: sidebar navigation polish, with improved active states, hover states, and spacing
v0.1.312 April 2026
- Dashboard: sidebar navigation now replaces the top header tabs
- Audit list: project name, date and time, audit scale, and a Latest badge now appear on each entry
- New audit page: cleaner pill-style URL input to match the homepage
- Auth: magic links now sign you in on the original tab instead of opening duplicate windows
- Pricing: quick and standard audits are free with no account required
- Billing: usage-based pricing for standard and deep audits at 0.10 per 1,000 crawwwl AI output tokens
v0.1.22 April 2026
- Performance: improved detection of unoptimised images and missing lazy loading
- Scoring: resolved an edge case where categories with no applicable rules inflated the overall score
- crawwwl AI: response handling normalised for more consistent output
- crawwwl AI: placeholder insight lines no longer appear in deep audit results
- CLI: terminal output polished and score bars align better on narrow terminals
- CLI: fix-check now exits with code 1 when critical findings are present
v0.1.118 March 2026
- Accessibility: added detection for empty buttons and icon-only links without ARIA labels
- SEO: improved Open Graph tag detection for Next.js 14+ metadata
- Security: reduced false positives for eval() usage inside comments and strings
- CLI: cleaner
--markdownoutput for sharing reports in GitHub issues - API: audit status polling now returns partial progress while a run is still processing
- crawwwl AI: fallback warning logs now report correctly when AI is unavailable
v0.1.04 March 2026
- Platform: initial release of crawwwl
- Scanners: five scanner categories: performance, security, code quality, SEO, and accessibility
- Audits: three audit scales: quick with no AI, standard with crawwwl AI, and deep with full reasoning
- CLI: audit, report, fix-check, and login commands
- Billing: usage-based pricing at 0.10 per 1,000 crawwwl AI output tokens, with a 5 minimum top-up and no subscription
- Scanners: support for
// crawwwl-ignoreon individual findings - Platform: Next.js and React project detection with context-aware scoring