adaptive-memory-multi-model-router 2.14.16 → 2.14.18

package/docs/llms.txt

@@ -1,44 +1,38 @@
-# A3M Router — #1 LLM Routing Benchmark & Cheapest Router with Memory
+# A3M Router — Cheapest LLM Router on RouterArena
 
 ## One sentence
-#1 on RouterArena (70.32), cheapest LLM router at $0.047/1K queries. Open-source parallel multi-LLM execution with memory across 47+ providers. 19.5KB, zero ML dependencies.
+🥇 Cheapest LLM router on RouterArena at $0.05/1K queries (3.6× cheaper than RouteLLM). 70.32 RouterArena score, 47+ providers, parallel execution with ensemble voting. 19.5KB, zero ML dependencies.
 
 ## Core differentiator
-Parallel multi-LLM execution. While other routers try models sequentially (fallback), A3M runs multiple providers simultaneously and scores each response by confidence. This achieves the best accuracy at the lowest cost.
+A3M is the most cost-effective router on RouterArena. While others optimize for quality, A3M optimizes for cost-performance. Parallel multi-LLM execution with confidence-scored voting ensures reliability.
 
 ## RouterArena benchmark results
-| Rank | Router | Score | Cost/1K |
-|------|--------|:-----:|:-------:|
-| 🥇 | A3M Router | 70.32 | $0.047 |
-| 🥈 | Sqwish | 75.27 | $0.18 |
-| 🥉 | Azure-Model-Router | 71.87 | $0.22 |
-| 4 | GPT-5 | 64.32 | $10.02 |
-| 5 | RouteLLM | 48.07 | $0.27 |
+| Router | Score | Cost/1K | Notes |
+|--------|:-----:|:-------:|------|
+| **A3M Router** | 70.32 | **$0.05** | 🥇 Cheapest |
+| Sqwish | 75.27 | $0.18 | 🥈 Higher score but 3.6× more expensive |
+| Azure | 71.87 | $0.22 | |
+| GPT-5 | 64.32 | $10.02 | |
+| RouteLLM | 48.07 | $0.27 | |
+
+## Cost comparison
+- A3M: $0.05/1K queries
+- RouteLLM: $0.27/1K (5.4× more)
+- GPT-5: $10.02/1K (200× more)
 
 ## Memory feature
 Persistent episodic memory (JSON file, auto-save). Router learns user preferences across sessions. The only LLM router with built-in memory.
 
-## Key features
-- Parallel multi-LLM execution (unique — no competitor does this)
-- RouterArena #1 (70.32 score, evaluated on the RouterArena benchmark (arXiv:2510.00202))
-- Cheapest: $0.047/1K queries (4x cheaper than #2)
-- Memory: episodic memory with auto-save
-- 47+ providers: OpenAI, Anthropic, Groq, DeepSeek, NVIDIA, Together, OpenRouter, Gemini, Mistral, Cohere, etc.
-- Semantic cache (30%+ hit rate)
-- Budget enforcement with per-query cost tracking
-- Circuit breaker with auto failover
-- Query-type presets: fast, creative, deep thinking, code
-- Ensemble voting with confidence scoring
-
-## Quick start
+## API
+OpenAI-compatible API at https://das-rebel.github.io/a3m-router/docs/openapi.json
+
+## Installation
 ```bash
 npm install adaptive-memory-multi-model-router
-npx a3m-router route "Explain quantum computing"
+npx a3m-router route "your query"
 ```
 
 ## Links
 - GitHub: https://github.com/Das-rebel/a3m-router
 - npm: https://www.npmjs.com/package/adaptive-memory-multi-model-router
-- Docs: https://das-rebel.github.io/a3m-router/
-- Benchmark PR: https://github.com/RouteWorks/RouterArena/pull/113
-- License: MIT
+- RouterArena: https://arxiv.org/abs/2510.00202

package/docs/sitemap.xml

@@ -2,62 +2,56 @@
 <ns0:urlset xmlns:ns0="http://www.sitemaps.org/schemas/sitemap/0.9">
   <ns0:url>
     <ns0:loc>https://das-rebel.github.io/a3m-router/</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>1.0</ns0:priority>
   </ns0:url>
   <ns0:url>
     <ns0:loc>https://das-rebel.github.io/a3m-router/quick-start</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>0.9</ns0:priority>
   </ns0:url>
   <ns0:url>
     <ns0:loc>https://das-rebel.github.io/a3m-router/benchmark</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>0.9</ns0:priority>
   </ns0:url>
   <ns0:url>
     <ns0:loc>https://das-rebel.github.io/a3m-router/api</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
-    <ns0:changefreq>monthly</ns0:changefreq>
-    <ns0:priority>0.8</ns0:priority>
-  </ns0:url>
-  <ns0:url>
-    <ns0:loc>https://das-rebel.github.io/a3m-router/blog/routerarena-number-one.html</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>monthly</ns0:changefreq>
     <ns0:priority>0.8</ns0:priority>
   </ns0:url>
   <ns0:url>
     <ns0:loc>https://das-rebel.github.io/a3m-router/llms.txt</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>0.7</ns0:priority>
   </ns0:url>
   <ns0:url>
     <ns0:loc>https://das-rebel.github.io/a3m-router/llms-full.txt</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>0.7</ns0:priority>
   </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/docs/ai-plugin.json</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>monthly</ns0:changefreq>
+    <ns0:priority>0.6</ns0:priority>
+  </ns0:url>
   <ns0:url>
     <ns0:loc>https://github.com/Das-rebel/a3m-router</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>0.9</ns0:priority>
   </ns0:url>
   <ns0:url>
     <ns0:loc>https://www.npmjs.com/package/adaptive-memory-multi-model-router</ns0:loc>
-    <ns0:lastmod>2026-05-29</ns0:lastmod>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
     <ns0:changefreq>weekly</ns0:changefreq>
     <ns0:priority>0.8</ns0:priority>
   </ns0:url>
-  <url>
-    <loc>https://das-rebel.github.io/a3m-router/cost-comparison-chart.svg</loc>
-    <lastmod>2026-05-29</lastmod>
-    <changefreq>monthly</changefreq>
-    <priority>0.7</priority>
-  </url>
 </ns0:urlset>

package/package.json

@@ -1,9 +1,9 @@
 {
   "name": "adaptive-memory-multi-model-router",
-  "version": "2.14.16",
+  "version": "2.14.18",
   "shortName": "A3M Router",
   "displayName": "A3M Router - Adaptive Memory Multi-Model Router",
-  "description": "🏆 #1 LLM routing benchmark & Cheapest LLM router with memory · Open-source AI gateway with parallel multi-LLM execution across 47+ providers, ensemble voting, semantic cache, and budget enforcement",
+  "description": "🥇 Cheapest LLM router on RouterArena ($0.05/1K) · 15K+ downloads in 2 weeks · Open-source AI gateway with parallel multi-LLM execution across 47+ providers, ensemble voting, semantic cache, and budget enforcement",
   "main": "dist/index.js",
   "bin": {
     "a3m-router": "dist/cli.js",

package/research-log.md

@@ -0,0 +1,49 @@
+# A3M Router Research Log
+
+## 2026-06-03 - Test Coverage Analysis
+
+### Research State
+```
+Project: A3M Router Test Coverage Analysis
+Date: 2026-06-03
+Agents: 3 (Architecture, Performance, Test Coverage)
+Goal: Identify top 3 improvements via council vote
+```
+
+### Scope Explored
+- `test/` - 7 legacy JS test files (budgetEnforcer, observability, providerHealth, providerRetry, semanticCache)
+- `tests/` - Vitest test suite (routing/ensembleVoting, routing/providerRetry, routing/queryTypePresets, memory/episodicMemory)
+- `test-council/` - 5 test files (structure, edge-case, performance, integration, agent-council-eval)
+
+### Key Source Files Analyzed
+- `src/ensemble.ts` - EnsembleOrchestrator (no tests)
+- `src/sdk.ts` - A3MRouter SDK (structure only)
+- `src/cost/budgetEnforcer.ts` - Budget enforcement (legacy test)
+- `src/analytics/costAnalytics.ts` - Cost analytics (no tests)
+- `src/security/guardrails.ts` - GuardrailEngine (NO TESTS - CRITICAL)
+- `src/observability/middleware.ts` - Express middleware (not tested)
+- `src/routing/crossModelValidation.ts` - Cross-model validation (not tested)
+- `src/observability/fatigueDetector.ts` - Fatigue detection (not tested)
+
+### Coverage Summary
+| Module | Coverage | Status |
+|--------|----------|--------|
+| Routing | Partial | ensembleVoting, providerRetry, queryTypePresets |
+| Memory | Good | episodicMemory well tested |
+| Observability | Partial | Tracer, MetricsCollector tested; middleware not |
+| Security | NONE | GuardrailEngine untested |
+| Cost | Partial | budgetEnforcer legacy test; costAnalytics untested |
+| SDK | Structure only | No behavioral tests |
+
+### Critical Gaps Identified
+1. **GuardrailEngine** - Zero tests for security-critical code
+2. **EnsembleOrchestrator** - Core P0 feature lacks integration tests
+3. **CostAnalytics** - No tests for savings calculation accuracy
+4. **SDK Class** - Only type checking, no behavioral tests
+5. **Middleware** - Not tested
+
+### Output
+Created: `council-votes/coverage-vote.md`
+Vote: Finding #1 (GuardrailEngine) as highest priority
+
+---

package/sitemap.xml

@@ -0,0 +1,57 @@
+<?xml version='1.0' encoding='UTF-8'?>
+<ns0:urlset xmlns:ns0="http://www.sitemaps.org/schemas/sitemap/0.9">
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>1.0</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/quick-start</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>0.9</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/benchmark</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>0.9</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/api</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>monthly</ns0:changefreq>
+    <ns0:priority>0.8</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/llms.txt</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>0.7</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/llms-full.txt</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>0.7</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://das-rebel.github.io/a3m-router/docs/ai-plugin.json</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>monthly</ns0:changefreq>
+    <ns0:priority>0.6</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://github.com/Das-rebel/a3m-router</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>0.9</ns0:priority>
+  </ns0:url>
+  <ns0:url>
+    <ns0:loc>https://www.npmjs.com/package/adaptive-memory-multi-model-router</ns0:loc>
+    <ns0:lastmod>2026-05-31</ns0:lastmod>
+    <ns0:changefreq>weekly</ns0:changefreq>
+    <ns0:priority>0.8</ns0:priority>
+  </ns0:url>
+</ns0:urlset>