Drop any log file and instantly get a full 12-tab dashboard — error breakdown, activity timeline, AI-crawler & GEO/AEO visibility, security & anomaly detection, traffic sources, devices, geography, performance, sessions and a 404 report — then export a client-ready report in one click.
New to log analysis? Click any question below — every chart in the dashboard above also has its own short caption explaining what it shows.
LogCraft parses your file client-side and runs it through fifteen analysis modules — the file never leaves your browser, so it's safe to drop in logs with internal hostnames, IPs or user data.
Auto-detects any Combined/Common Log Format variant (Apache/Nginx and countless custom fields), structured JSON-lines logs, syslog, and generic leveled app logs — no config, no format selection needed.
Counts ERROR / WARN / INFO / DEBUG lines, computes error rate, and flags when it's outside a healthy range.
Buckets events over time, adds an hour-of-day pattern view, and compares the first vs. second half of the log to show if things are improving or worsening.
HTTP status breakdown, top client IPs, top endpoints, and a dedicated 404/broken-link report showing exactly which URLs need fixing or redirecting.
Identifies every bot — including GPTBot, ClaudeBot, PerplexityBot and Google-Extended — shows which pages AI engines are reading, per-bot crawl efficiency, and a robots.txt/llms.txt check.
Flags IPs showing path-traversal/SQLi/XSS probing, scripted (non-browser) clients, brute-force auth failures, or abnormal request bursts — with a risk-ranked table.
Splits human traffic into AI-assistant referrals, search engines, social and direct — plus the exact landing pages AI-referred visitors clicked into.
Desktop/mobile/tablet split, browser and OS breakdown, parsed straight from user-agent strings.
Breaks down requests by country when location data is present in the log.
Average/median/P95 response time and the slowest endpoints, when response-time data is present.
Session count, pages per session and a bounce-style single-page-session rate, when session data is present.
Groups similar error messages together so you can see which single issue is costing you the most.
Turns every module's findings into plain-language, prioritised next steps — critical, warning, and informational.
Every table exports to CSV individually; the full report exports as a shareable TXT or a chart-included PDF.
Click any column to sort, filter tables by keyword, and share a direct link straight to any tab.
Drag a .log or .txt file onto the drop zone (or click to browse), optionally add a keyword to highlight in the raw log view, and click Analyse Log File. Parsing happens instantly in your browser — nothing is sent to a server. Extra tabs like Bots & AI Crawlers, Security & Anomalies, Traffic Sources, Devices, Geography, Performance and Sessions appear automatically whenever your log contains that data. Use Export TXT or Export PDF for the full report, or the Export CSV button on any individual table.
The Log Health Score (0–100) weighs error rate, server errors, security signals, bot/AI-crawler behaviour and how concentrated your failures are around a single root cause. 75+ = Healthy · 50–74 = Needs Attention · Below 50 = Investigate Now. Fix red (critical) items first, then amber warnings.