AI search engine and MCP server for bots and agents. One Rust binary indexes code, docs, logs, JSONL/CSV and PDFs for BM25 full-text, vector and hybrid search, code search without grep, agent memory and RAG. Elasticsearch-compatible API; Corpus Hub serves hundreds of reference memory corpora.
Rust agent-memoryai-agentsbm25claude-codeclicode-searchcursordatabaseelasticsearchfull-text-searchhybrid-searchjsonllocal-firstmcpmcp-servermodel-context-protocolragrustsearch-enginevector-search
site(use-cases): split cybersecurity into two separate case-study cards
xerj-orgc778ccda31b9
diff --git a/landing/use-cases.html b/landing/use-cases.htmlindex bbb85b122..8437f106e 100644--- a/landing/use-cases.html+++ b/landing/use-cases.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html" class="active">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html">AI SEARCH</a>@@ -63,15 +64,67 @@</p></section>-<!-- 01 AI SECURITY REVIEW — the real, full-story flagship -->+<!-- ═══ GROUP: CYBERSECURITY ═══ -->+<section class="scene-section" style="border-top:0; padding-bottom:0;">+ <div class="kicker"><span>GROUP</span><span class="dash">·</span><span>CYBERSECURITY</span></div>+ <h2 class="scene" style="font-size:var(--fs-56);">AUDIT REAL CODE. <span class="accent">FIND REAL BUGS.</span></h2>+ <p class="wide">Two reproducible whitebox security audits with XERJ as the agent's retrieval substrate — one on a large third-party codebase, one on XERJ's own engine. Both real, both measured, each its own full case study.</p>+</section>++<!-- 01 CYBERSECURITY · WORDPRESS SECURITY AUDIT --><div class="uc-showcase"><div class="uc-copy"><span class="uc-num">01</span>- <div class="uc-title">FIND REAL BUGS.<br>INDEX ONCE,<br><span class="accent">QUERY THE FLAW.</span></div>- <p class="uc-pitch">The flagship isn't a benchmark — it's an audit that found and fixed real vulnerabilities. XERJ indexed its own Rust engine (197 files) as security <em>facts</em> and a call graph, then a query grep cannot express surfaced an unauthenticated <code class="inline">/_sql</code> stack-overflow that abends the process, plus a search-template injection that leaks private documents and a snapshot path-traversal — <strong>all fixed on the branch, proven by crashing then not-crashing a live server</strong>. The same method on real WordPress core audited 1,492 files for ~26K tokens. Two codebases, real findings, every number reproducible.</p>- <div class="uc-metric">6/6 KNOWN BUGS RECALLED · 171× FEWER TOKENS · 4 NEW HIGHS, 3 FIXED</div>- <div class="uc-persona">CISO · APPSEC ENGINEER · SECURITY RESEARCHER · AI-AGENT BUILDER</div>- <a href="/use-cases/code-security-audit.html" class="uc-cta">FULL USE CASE →</a>+ <div class="uc-title">WORDPRESS<br>SECURITY<br><span class="accent">AUDIT.</span></div>+ <p class="uc-pitch">An AI agent security-audited real WordPress core — 1,492 PHP files, ~619k lines — using XERJ as its retrieval substrate: index once as security facts (11,990 functions, 1,343 hooks), query the bug shape, read only the survivors. It swept 226 handlers for missing-capability IDOR (0 found — core is hardened) and surfaced the real <code class="inline">wp_http_validate_url</code> SSRF gap (169.254 cloud-metadata) reachable unauthenticated. The point is the method and its economics: the whole audit cost ~26K tokens against ~5.2M just to load the tree.</p>+ <div class="uc-metric">~26K-TOKEN AUDIT vs ~5.2M TO LOAD · 199× FEWER</div>+ <div class="uc-persona">CISO · APPSEC ENGINEER · SECURITY RESEARCHER</div>+ <a href="/case-studies/wordpress-security-audit.html" class="uc-cta">FULL CASE STUDY →</a>+ </div>+ <div class="uc-viz">+ <div class="kicker"><span>MEASURED · WORDPRESS CORE · 1,492 FILES / ~619K LINES</span></div>+ <div class="cmp-bars">+ <div class="cmp-bars-head">+ <span></span>+ <span class="cmp-a">XERJ-ASSISTED</span>+ <span class="cmp-b">READ EVERY FILE</span>+ </div>+ <div class="cmp-bars-row">+ <span class="cmp-label">Tokens, entire audit</span>+ <span class="cmp-val cmp-a">+ <span class="cmp-bar" style="width:0.45%;"></span>+ <span class="cmp-num">~26,000</span>+ </span>+ <span class="cmp-val cmp-b">+ <span class="cmp-bar" style="width:90.0%;"></span>+ <span class="cmp-num">~5,200,000 <span class="cmp-ratio">199×</span></span>+ </span>+ </div>+ <div class="cmp-bars-row">+ <span class="cmp-label">SQLi triage: must read after</span>+ <span class="cmp-val cmp-a">+ <span class="cmp-bar" style="width:7.1%;"></span>+ <span class="cmp-num">4 functions</span>+ </span>+ <span class="cmp-val cmp-b">+ <span class="cmp-bar" style="width:90.0%;"></span>+ <span class="cmp-num">51 whole files</span>+ </span>+ </div>+ </div>+ <p style="margin-top:14px; font-size:12px; letter-spacing:.04em; color:var(--z-mute); font-family:var(--font-data);">TOKEN COST TO THE ANSWER · NOT LATENCY · CORE IS HARDENED — 0 IDOR ACROSS 226 HANDLERS; THE ONE REAL WEAKNESS (SSRF) IS STATED PLAINLY · SOURCES IN <a href="https://github.com/xerj-org/xerj/tree/main/docs/case-studies/wordpress-security-audit" rel="external">docs/case-studies/wordpress-security-audit →</a></p>+ </div>+</div>++<!-- 02 CYBERSECURITY · XERJ SELF-AUDIT -->+<div class="uc-showcase">+ <div class="uc-copy">+ <span class="uc-num">02</span>+ <div class="uc-title">XERJ AUDITS<br>ITS OWN<br><span class="accent">RUST.</span></div>+ <p class="uc-pitch">The stronger proof: XERJ turned the method on its own Rust engine (197 files) and found and <em>fixed</em> real vulnerabilities. A call-graph query grep cannot express surfaced an unauthenticated <code class="inline">/_sql</code> stack-overflow that aborts the process; a five-stage instruction/data pass added a search-template injection that leaks private documents and a snapshot path-traversal. The three application-layer highs are fixed on the branch, each proven by crashing — then not-crashing — a live server. Against six known-fixed bugs the substrate recalled 6/6 where grep managed 4/6, at 171× fewer tokens.</p>+ <div class="uc-metric">4 REAL HIGHS · 3 FIXED · 6/6 RECALL vs grep 4/6 · 171× FEWER TOKENS</div>+ <div class="uc-persona">APPSEC ENGINEER · SECURITY RESEARCHER · AI-AGENT BUILDER · RUST TEAM</div>+ <a href="/case-studies/xerj-self-audit.html" class="uc-cta">FULL CASE STUDY →</a></div><div class="uc-viz"><div class="kicker"><span>MEASURED · tiktoken cl100k_base · TOKENS TO REACH THE /_sql FINDING</span></div>@@ -104,14 +157,21 @@</span></div></div>- <p style="margin-top:14px; font-size:12px; letter-spacing:.04em; color:var(--z-mute); font-family:var(--font-data);">TOKEN COST TO THE ANSWER · NOT LATENCY · RECALL MEASURED AGAINST 6 KNOWN-FIXED BUGS (grep CAN'T FIND MISSING CODE) · FULL METHODOLOGY, ALL 8 FINDINGS & THE IN-SAMPLE CAVEAT IN <a href="https://github.com/xerj-org/xerj/tree/main/docs/case-studies/xerj-self-audit" rel="external">docs/case-studies/xerj-self-audit →</a> · WORDPRESS PROOF IN <a href="https://github.com/xerj-org/xerj/tree/main/docs/case-studies/wordpress-security-audit" rel="external">wordpress-security-audit →</a></p>+ <p style="margin-top:14px; font-size:12px; letter-spacing:.04em; color:var(--z-mute); font-family:var(--font-data);">TOKEN COST TO THE ANSWER · NOT LATENCY · RECALL MEASURED AGAINST 6 KNOWN-FIXED BUGS (grep CAN'T FIND MISSING CODE); 6/6 IS IN-SAMPLE — SEE THE CAVEAT · METHODOLOGY + ALL 8 FINDINGS IN <a href="https://github.com/xerj-org/xerj/tree/main/docs/case-studies/xerj-self-audit" rel="external">docs/case-studies/xerj-self-audit →</a></p></div></div>-<!-- 02 SIEM — the headline chart IS the comparison -->+<!-- ═══ GROUP: PERFORMANCE & SCALE vs ELASTICSEARCH ═══ -->+<section class="scene-section" style="border-top:0; padding-bottom:0;">+ <div class="kicker"><span>GROUP</span><span class="dash">·</span><span>PERFORMANCE & SCALE</span></div>+ <h2 class="scene" style="font-size:var(--fs-56);">MEASURED <span class="accent">vs ELASTICSEARCH 8.13.</span></h2>+ <p class="wide">Reproducible capability benchmarks from <code class="inline">demo/playbooks/SCORECARD.md</code> — measured numbers on a real query matrix, not user stories.</p>+</section>++<!-- 03 SIEM — the headline chart IS the comparison --><div class="uc-showcase"><div class="uc-copy">- <span class="uc-num">02</span>+ <span class="uc-num">03</span><div class="uc-title">SECURITY<br>ANALYTICS<br><span class="accent">AT SCALE.</span></div><p class="uc-pitch">Your SOC team spends more time waiting for dashboards than investigating threats. The battery on the right is the reproducible XERJ-vs-Elasticsearch query matrix — bool filters, range scans, term lookups and aggregations — measured on a 1M-doc corpus. XERJ is the yellow bar; Elasticsearch 8.13 is the gray one.</p><div class="uc-metric">11.5× FASTER BOOL QUERIES</div>@@ -212,10 +272,10 @@</div></div>-<!-- 03 LOGS — ingest throughput over 24h -->+<!-- 04 LOGS — ingest throughput over 24h --><div class="uc-showcase"><div class="uc-copy">- <span class="uc-num">03</span>+ <span class="uc-num">04</span><div class="uc-title">OPERATIONAL<br><span class="accent">INTELLIGENCE.</span></div><p class="uc-pitch">Ship logs at line rate. Query hot data in milliseconds. No ILM policy puzzles, no shard tuning, no Logstash pipeline. The chart shows a 24-hour diurnal ingest pattern — XERJ sustains it with zero write-stalls, and ingests 1.5–1.7× faster than Elasticsearch in the reproducible benchmark.</p><div class="uc-metric">1.7× FASTER INGEST</div>@@ -238,10 +298,10 @@</div></div>-<!-- 04 AI — intent clusters + memory comparison -->+<!-- 05 AI — intent clusters + memory comparison --><div class="uc-showcase"><div class="uc-copy">- <span class="uc-num">04</span>+ <span class="uc-num">05</span><div class="uc-title">AI-NATIVE<br>SEARCH &<br><span class="accent">RETRIEVAL.</span></div><p class="uc-pitch">Hybrid semantic + keyword search in one query. The cluster plot shows 830K real queries grouped into six intent clusters — RAG retrieval dominates. Every cluster is one index query away, not a separate Pinecone call. The benchmark below is the reproducible XERJ-vs-Elasticsearch latency for kNN and text retrieval.</p><div class="uc-metric">3.4× FASTER kNN · 100% RECALL</div>@@ -302,10 +362,10 @@</div></div>-<!-- 05 ES REPLACEMENT — resource comparison is THE chart -->+<!-- 06 ES REPLACEMENT — resource comparison is THE chart --><div class="uc-showcase"><div class="uc-copy">- <span class="uc-num">05</span>+ <span class="uc-num">06</span><div class="uc-title">ELASTICSEARCH<br><span class="accent">REPLACEMENT.</span></div><p class="uc-pitch">Same API on port 9200. Same query DSL. Same client libraries. The comparison on the right is the reproducible query-latency head-to-head. Yellow is XERJ; gray is Elasticsearch 8.13 on the identical workload.</p><div class="uc-metric">1.20× SMALLER ON DISK</div>@@ -386,10 +446,10 @@</div></div>-<!-- 06 OBSERVABILITY — stack comparison -->+<!-- 07 OBSERVABILITY — stack comparison --><div class="uc-showcase"><div class="uc-copy">- <span class="uc-num">06</span>+ <span class="uc-num">07</span><div class="uc-title">UNIFIED<br><span class="accent">OBSERVABILITY.</span></div><p class="uc-pitch">Logs, traces, and metrics in one store — one binary instead of a Grafana-Loki-Tempo-Mimir stack to babysit. OTLP in, Prometheus out. The chart on the right is the reproducible query-latency head-to-head vs Elasticsearch 8.13 on log-shaped data.</p><div class="uc-metric">6 SYSTEMS → 1 BINARY</div>diff --git a/landing/use-cases/ai-search-retrieval.html b/landing/use-cases/ai-search-retrieval.htmlindex 24f8a000c..7b618a268 100644--- a/landing/use-cases/ai-search-retrieval.html+++ b/landing/use-cases/ai-search-retrieval.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html" class="active">AI SEARCH</a>diff --git a/landing/use-cases/code-security-audit.html b/landing/use-cases/code-security-audit.htmlindex a0c1ed46f..93e458b47 100644--- a/landing/use-cases/code-security-audit.html+++ b/landing/use-cases/code-security-audit.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html" class="active">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html">AI SEARCH</a>diff --git a/landing/use-cases/elasticsearch-replacement.html b/landing/use-cases/elasticsearch-replacement.htmlindex 50866df43..79913e7fc 100644--- a/landing/use-cases/elasticsearch-replacement.html+++ b/landing/use-cases/elasticsearch-replacement.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html">AI SEARCH</a>diff --git a/landing/use-cases/operational-intelligence.html b/landing/use-cases/operational-intelligence.htmlindex 385d19679..1686bad46 100644--- a/landing/use-cases/operational-intelligence.html+++ b/landing/use-cases/operational-intelligence.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html" class="active">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html">AI SEARCH</a>diff --git a/landing/use-cases/security-analytics.html b/landing/use-cases/security-analytics.htmlindex d5d35371a..5ecb094d8 100644--- a/landing/use-cases/security-analytics.html+++ b/landing/use-cases/security-analytics.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html" class="active">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html">AI SEARCH</a>diff --git a/landing/use-cases/unified-observability.html b/landing/use-cases/unified-observability.htmlindex d10be768f..8a066bed1 100644--- a/landing/use-cases/unified-observability.html+++ b/landing/use-cases/unified-observability.html@@ -40,7 +40,8 @@<nav class="uc-subnav" aria-label="Use cases"><a href="/use-cases.html">ALL USE CASES</a>- <a href="/use-cases/code-security-audit.html">AI SECURITY REVIEW</a>+ <a href="/case-studies/wordpress-security-audit.html">WORDPRESS AUDIT</a>+ <a href="/case-studies/xerj-self-audit.html">XERJ SELF-AUDIT</a><a href="/use-cases/security-analytics.html">SECURITY ANALYTICS</a><a href="/use-cases/operational-intelligence.html">OPERATIONAL INTELLIGENCE</a><a href="/use-cases/ai-search-retrieval.html">AI SEARCH</a>