Roadmap¶
AgentTel follows a phased release plan, building from core instrumentation through GenAI support to a complete agent interface layer.
Release History¶
v0.1.0-alpha — Foundation + GenAI + Agent Layer¶
Status: Implemented
Phase 1: Foundation¶
Core instrumentation and span enrichment.
- [x] Gradle multi-module project setup (Java 17+)
- [x]
agenttel-api— Annotations (@AgentOperation,@AgentObservable,@DeclareDependency), attribute constants, enums - [x]
agenttel-core—AgentTelSpanProcessorwith topology, baseline, and decision enrichment - [x]
agenttel-core—TopologyRegistrywith dependency graph serialization - [x]
agenttel-core—StaticBaselineProviderfrom annotation metadata - [x]
agenttel-core—RollingBaselineProviderwith lock-free ring buffer sliding windows - [x]
agenttel-core—CompositeBaselineProviderchaining static → rolling → default - [x]
agenttel-core—AnomalyDetectorwith z-score based deviation detection - [x]
agenttel-core—PatternMatcherdetecting cascade failures, latency degradation, error rate spikes, memory leaks - [x]
agenttel-core—SloTrackerwith error budget consumption and threshold alerting - [x]
agenttel-core—AgentTelEventEmitterfor structured events via OTel Logs API - [x]
agenttel-spring-boot-starter— Auto-configuration with@ConfigurationProperties - [x]
agenttel-testing— Test utilities - [x] CI/CD with GitHub Actions (JDK 17 + 21 matrix)
- [x] Grafana dashboard templates (overview + GenAI)
Phase 2: GenAI Instrumentation¶
Observability for AI/ML workloads on the JVM.
- [x]
agenttel-genai—gen_ai.*andagenttel.genai.*attribute constants - [x]
agenttel-genai—ModelCostCalculatorwith per-model pricing for Claude, GPT, and Bedrock models - [x]
agenttel-genai—SpringAiSpanEnricherenriching existing Micrometer spans - [x]
agenttel-genai—CostEnrichingSpanExporterwith delegatingSpanDatapattern - [x]
agenttel-genai— LangChain4j instrumentation:TracingChatLanguageModel,TracingStreamingChatLanguageModel,TracingEmbeddingModel,TracingContentRetriever - [x]
agenttel-genai— Anthropic SDK instrumentation:TracingAnthropicClient - [x]
agenttel-genai— OpenAI SDK instrumentation:TracingOpenAIClient - [x]
agenttel-genai— AWS Bedrock instrumentation:TracingBedrockRuntimeClient - [x] Spring Boot auto-configuration with
@ConditionalOnClassdetection - [x] LangChain4j example application
Phase 3: Agent Interface Layer¶
Complete toolkit for AI agent interaction with production systems.
- [x]
agenttel-agent—ServiceHealthAggregatorfor real-time health from span data - [x]
agenttel-agent—IncidentContext/IncidentContextBuilderfor structured incident packages - [x]
agenttel-agent—RemediationAction/RemediationRegistry/RemediationExecutorfor agent-executable fixes - [x]
agenttel-agent—AgentActionTrackerrecording agent decisions as OTel spans - [x]
agenttel-agent—ContextFormatterwith prompt-optimized output formats (compact, full, JSON) - [x]
agenttel-agent—AgentContextProvideras single entry point for agent queries - [x]
agenttel-agent—McpServerimplementing Model Context Protocol over JSON-RPC - [x]
agenttel-agent—AgentTelMcpServerBuilderwith 5 pre-registered tools - [x] Full test coverage for all agent layer components
Phase 4: Frontend Telemetry & Tooling¶
Full-stack telemetry and developer experience.
- [x]
agenttel-web— TypeScript browser SDK (@agenttel/web) with auto-instrumentation of page loads, SPA navigation, API calls, clicks, and errors - [x]
agenttel-web— Journey tracking with multi-step funnel detection, completion rates, and abandonment analysis - [x]
agenttel-web— Client-side anomaly detection: rage clicks, API failure cascades, slow page loads, error loops, funnel drop-offs - [x]
agenttel-web— W3C Trace Context injection and backend trace ID correlation for cross-stack linking - [x]
agenttel-web— Per-route baselines and decision metadata (escalation levels, runbooks, fallbacks) - [x]
agenttel-web— React checkout example application - [x]
agenttel-instrument— Python MCP server for IDE-based instrumentation automation - [x]
agenttel-instrument—analyze_codebasetool for Java/Spring Boot source scanning - [x]
agenttel-instrument—instrument_backendandinstrument_frontendconfig generation - [x]
agenttel-instrument—validate_instrumentationwith source code cross-referencing - [x]
agenttel-instrument—suggest_improvementsandapply_improvementswith live health data integration - [x]
agenttel-agent—TrendAnalyzerfor operation metric trend analysis - [x]
agenttel-agent—SloReportGeneratorfor LLM-optimized SLO compliance reports - [x]
agenttel-agent—ExecutiveSummaryBuilderfor high-level service status summaries - [x]
agenttel-agent—CrossStackContextBuilderfor correlated frontend-backend context
Planned: v0.2.0-alpha¶
Agent Layer Enhancements¶
- [ ] MCP Server authentication (API key, mTLS)
- [ ] Server-Sent Events (SSE) transport for real-time streaming
- [ ] WebSocket transport option
- [ ] Configurable tool permissions per agent identity
- [ ] Rate limiting on MCP tool invocations
- [ ] Remediation action execution hooks (webhook, Kubernetes API, AWS API)
Observability Enhancements¶
- [ ] Causality tracking — linking anomalies to recent deployments and config changes
- [ ] Dependency health state machine with transition events
- [ ] Multi-service correlation — cross-service incident grouping
- [ ] Custom metric emission alongside span enrichment
GenAI Enhancements¶
- [ ] Prompt/response content capture (opt-in, with PII redaction)
- [ ] Token budget tracking and alerting
- [ ] Model performance comparison spans
- [ ] Guardrail integration for content safety monitoring
Planned: v0.3.0-beta¶
Framework Ecosystem¶
- [ ] Quarkus extension (
agenttel-quarkus) - [ ] Micronaut module (
agenttel-micronaut) - [ ] Ktor plugin (
agenttel-ktor) - [ ] Kotlin DSL for idiomatic Kotlin configuration
Infrastructure Integration¶
- [ ] OTel Collector processor (Go) for server-side enrichment and validation
- [ ] Kubernetes operator for automatic MCP server deployment
- [ ] Helm chart for MCP server sidecar pattern
Testing and Validation¶
- [ ] Performance benchmarks (enrichment overhead per span)
- [ ] Load testing with sustained high throughput
- [ ] Security audit of MCP server
Planned: v1.0.0¶
Stability¶
- [ ] API stability guarantee — no breaking changes without major version bump
- [ ] Semantic versioning commitment
- [ ] Long-term support (LTS) release
Standards¶
- [ ] OpenTelemetry community proposal for
agenttel.*conventions - [ ] OTel GenAI SIG collaboration on
gen_ai.*attribute stabilization - [ ] CNCF project submission
Documentation¶
- [ ] Interactive tutorial / playground
- [ ] Video walkthroughs
- [ ] Migration guides from existing instrumentation
- [ ] Production deployment guide with reference architectures
Ecosystem¶
- [ ] Grafana plugin for AgentTel-specific visualizations
- [ ] VS Code extension for
@AgentOperationannotation assistance - [x] ~~CLI tool for validating AgentTel configuration~~ — implemented as
agenttel-instrumentMCP server withvalidate_instrumentationandsuggest_improvementstools
Compatibility Matrix¶
Backend (JVM)
| AgentTel Version | Java | OTel SDK | Spring Boot | LangChain4j | Spring AI |
|---|---|---|---|---|---|
| 0.1.0-alpha | 17, 21 | 1.59.0+ | 3.4.x | 1.0.0+ | 1.0.0+ |
| 0.2.0-alpha (planned) | 17, 21 | 1.60.0+ | 3.4.x, 3.5.x | 1.0.0+ | 1.0.0+ |
| 0.3.0-beta (planned) | 17, 21, 25 | 1.62.0+ | 3.x | 1.x | 1.x |
| 1.0.0 (planned) | 17+ | 1.x | 3.x | 1.x | 1.x |
Frontend (Browser)
| AgentTel Version | TypeScript | Browsers |
|---|---|---|
| 0.1.0-alpha | 4.7+ | Chrome, Firefox, Safari, Edge (ES2020+) |
Tooling
| AgentTel Version | Python (instrument agent) |
|---|---|
| 0.1.0-alpha | 3.11+ |