Skip to main content
Current Version: v1.4.0 — Observe Mode & Governance Dashboard (July 2026) Previous: v1.3.0 — Autonomous Agent Governance (May 2026) Next: v1.5.0 — Enterprise Platform (Q3 2026)

​
Release Timeline

​
v1.3.0 — Autonomous Agent Governance ✅

Status: Built — Launching May 2026 Theme: Identity governance, cryptographic evidence, temporal controls, platform adapters

​
10 Governance Domains

v1.3 expands TealTiger from 6 to 10 governance domains:

​
Key Features

FREEZE Rules

Immutable, non-overridable safety controls that persist across policy hot-swaps and log tamper attempts.

Non-Human Identity

Agent lifecycle management, scope enforcement, Zero Standing Privilege with JIT grants, cryptographic attestation.

TealProof

Merkle-tree governance receipts with RFC 3161 timestamping. Standalone Verification SDK for third-party audit.

TealFlow

Declarative YAML governance workflows with job dependencies, org-level inheritance, and floor enforcement.

Platform Adapters

AWS Bedrock Agents, AWS AgentCore, Azure AI Agent Service — native governance integration.

OWASP Policy Pack

Pre-built zero-config policies covering all 10 OWASP Agentic Security risks.

​
Distribution

  • npm: [email protected]
  • PyPI: tealtiger==1.3.0
  • Docker: tealtigeradmin/tealtiger-typescript:1.3.0
  • 12 LLM providers (TypeScript + Python parity)

v1.3.0 Release Details

Full bundle manifest, features, and migration guide

​
v1.2.0 — Governance Bundle ✅

Status: Released — May 2026 Theme: Parallel evaluation, evidence contracts, Docker sidecar
  • TealEngineV12 — Parallel module evaluation, most-restrictive-wins merge
  • TealSecrets — 500+ secret detection patterns with confidence scoring
  • TealRegistry — Model/tool allowlisting with provenance verification
  • TealReliability — Retry budgets, circuit breakers, fallback chains
  • TealMemory — Memory governance (5 scopes, 4 classifications)
  • BundleExporter — SARIF v2.1.0, JUnit XML, JSON evidence export
  • Docker Sidecar — Language-agnostic HTTP governance API

v1.2.0 Release Details

Complete release scope and migration guide

​
v1.1.x — Foundation ✅

Status: Released — Mar/Apr 2026 Theme: Core engine, security guardrails, 12 providers
  • TealEngine — Deterministic policy evaluation
  • TealGuard — PII, prompt injection, content moderation
  • TealMonitor — Cost tracking and budget management
  • TealCircuit — Circuit breaker for failure isolation
  • TealAudit — Audit logging with correlation IDs
  • 12 LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, Cohere, Mistral)

​
v1.4.0 — Industry Platform (Q1 2027) 📋

Status: Planning Theme: Ecosystem maturity, advanced analytics, community governance

​
Planned Directions

  • Multi-agent governance — Cross-agent policy coordination, delegation chains
  • Governance analytics — Security posture scoring, trend analysis, benchmarks
  • Community policy marketplace — Shared policy packs, community-contributed detectors
  • Advanced ML classifiers — Fine-tuned models for domain-specific detection
  • Formal verification — Property-based testing integration for policy correctness
  • Additional platform adapters — Google Vertex AI, Anthropic MCP native, LangGraph
v1.4.0 features are in early planning. Priorities will be shaped by community feedback and real deployment needs.

​
v2.0.0 — Industry Standard (Q4 2027) 🔮

Status: Future Vision Theme: Standard recognition, optional platform, ecosystem leadership
  • OWASP project contribution and standards co-authorship
  • Optional centralized policy management (SaaS)
  • Academic partnerships and published research
  • 50+ framework integrations
  • Industry benchmark leadership
SDK-First Commitment: The core SDKs will always remain open source (Apache 2.0) and fully functional without any platform. Platform features are optional enhancements.

​
Strategic Principles

​
Feedback & Contributions

GitHub Discussions

Share feedback, use cases, and feature requests

Email Us

Direct feedback to the team
Last updated: May 2026. This roadmap is updated with each release. Features and timelines may change based on community feedback and real deployment needs.