chore: align Python packaging metadata (version, requires-python, counts, single-source)

.github/linters/markdown-link-check.json

@@ -52,7 +52,7 @@
     { "pattern": "genai\\.owasp\\.org" },
     { "pattern": "learn\\.microsoft\\.com" },
     { "pattern": "microsoft\\.com/en-us/security/blog" },
-    { "pattern": "dcbadge\\.limes\\.pink/api/server/vBg9SNN8" },
+    { "pattern": "dcbadge\\.limes\\.pink/api/server/7aVPCcVh" },
     { "pattern": "agent-governance-python/agent-lightning/pull" },
     { "pattern": "azure-container-apps/blob" },
     { "pattern": "^(?!https?://|mailto:|#)" }

README.md

@@ -270,9 +270,9 @@ All five language SDKs implement core governance (policy, identity, trust, audit
 See **[Language Package Matrix](docs/PACKAGE-FEATURE-MATRIX.md)** for detailed per-language coverage.
 
 <details>
-<summary><b>Python distributions (v4.0.0 — consolidated)</b></summary>
+<summary><b>Python distributions (v4.1.0 — consolidated)</b></summary>
 
-As of v4.0.0, 45 packages have been consolidated into 5 top-level distributions:
+As of v4.1.0, 45 packages have been consolidated into 5 top-level distributions:
 
 | Distribution | PyPI | What's included |
 |--------------|------|-----------------|

agent-governance-python/README.md

@@ -39,6 +39,6 @@ It is not for applications or dashboards, demos or examples, monorepo-only produ
 
 `agent-compliance/`, `agent-discovery/`, `agent-hypervisor/`, `agent-lightning/`, `agent-marketplace/`, `agent-mcp-governance/`, `agent-mesh/`, `agent-os/`, `agent-primitives/`, `agent-rag-governance/`, `agent-runtime/`, `agent-sandbox/`, `agent-sre/`, `agentmesh-integrations/`
 
-## Package Consolidation (v4.0.0 — Complete)
+## Package Consolidation (v4.1.0 — Complete)
 
-As of v4.0.0, 45 packages have been consolidated into 5 top-level distributions: `agent-governance-toolkit-core`, `agent-governance-toolkit-runtime`, `agent-governance-toolkit-sre`, `agent-governance-toolkit-cli`, and the `agent-governance-toolkit[full]` meta-package. See [issue #2482](https://github.com/microsoft/agent-governance-toolkit/issues/2482) for details. The consolidation plan, audit data, and migration guide are in `docs/package-consolidation/`. Previous package names remain installable as stub packages that redirect to the consolidated distributions.
+As of v4.1.0, 45 packages have been consolidated into 5 top-level distributions: `agent-governance-toolkit-core`, `agent-governance-toolkit-runtime`, `agent-governance-toolkit-sre`, `agent-governance-toolkit-cli`, and the `agent-governance-toolkit[full]` meta-package. See [issue #2482](https://github.com/microsoft/agent-governance-toolkit/issues/2482) for details. The consolidation plan, audit data, and migration guide are in `docs/package-consolidation/`. Previous package names remain installable as stub packages that redirect to the consolidated distributions.

agent-governance-python/agent-compliance/docs/analyst/fact-sheet.md

@@ -27,7 +27,7 @@
 - **Circuit breakers and cascade protection** — Prevent single-agent failures from cascading across multi-agent systems
 - **DID-based cryptographic agent identity** — Decentralized identifiers for verifiable, non-repudiable agent authentication
 - **OpenTelemetry-native observability** — Export governance telemetry as traces, metrics, and logs to any OTEL-compatible backend
-- **12+ framework adapters** — LangChain, CrewAI, AutoGen, Semantic Kernel, LlamaIndex, Haystack, OpenAI Agents SDK, Google ADK, MCP, A2A, and more
+- **21 integration packages** — LangChain, CrewAI, AutoGen, Semantic Kernel, LlamaIndex, Haystack, OpenAI Agents SDK, Google ADK, MCP, A2A, and more
 - **Behavioral anomaly detection** — ML-based detection of unusual agent behavior patterns
 - **REST API for language-agnostic integration** — Use from any language, not just Python
 

agent-governance-python/agent-compliance/pyproject.toml

@@ -8,7 +8,7 @@ version = "4.1.0"
 description = "Public Preview — Unified installer and runtime policy enforcement for the Agent Governance Toolkit"
 readme = "README.md"
 license = {text = "MIT"}
-requires-python = ">=3.9"
+requires-python = ">=3.11"
 authors = [
     {name = "Microsoft Corporation", email = "agentgovtoolkit@microsoft.com"},
 ]
@@ -24,8 +24,6 @@ classifiers = [
     "Development Status :: 4 - Beta",
     "Intended Audience :: Developers",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.9",
-    "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Topic :: Security",
@@ -39,31 +37,31 @@ dependencies = [
 ]
 
 [project.optional-dependencies]
-core = ["agent-governance-toolkit-core>=4.0.0,<5.0"]
-integrations = ["agent-governance-toolkit-integrations>=4.0.0,<5.0"]
-cli = ["agent-governance-toolkit-cli>=4.0.0,<5.0"]
-protocols = ["agent-governance-toolkit-protocols>=4.0.0,<5.0"]
+core = ["agent-governance-toolkit-core>=4.1.0,<5.0"]
+integrations = ["agent-governance-toolkit-integrations>=4.1.0,<5.0"]
+cli = ["agent-governance-toolkit-cli>=4.1.0,<5.0"]
+protocols = ["agent-governance-toolkit-protocols>=4.1.0,<5.0"]
 # Framework-specific extras (delegated to integrations package)
-langchain = ["agent-governance-toolkit-integrations[langchain]>=4.0.0,<5.0"]
-crewai = ["agent-governance-toolkit-integrations[crewai]>=4.0.0,<5.0"]
-openai-agents = ["agent-governance-toolkit-integrations[openai-agents]>=4.0.0,<5.0"]
-langgraph = ["agent-governance-toolkit-integrations[langgraph]>=4.0.0,<5.0"]
-llamaindex = ["agent-governance-toolkit-integrations[llamaindex]>=4.0.0,<5.0"]
-haystack = ["agent-governance-toolkit-integrations[haystack]>=4.0.0,<5.0"]
-pydantic-ai = ["agent-governance-toolkit-integrations[pydantic-ai]>=4.0.0,<5.0"]
-adk = ["agent-governance-toolkit-integrations[adk]>=4.0.0,<5.0"]
+langchain = ["agent-governance-toolkit-integrations[langchain]>=4.1.0,<5.0"]
+crewai = ["agent-governance-toolkit-integrations[crewai]>=4.1.0,<5.0"]
+openai-agents = ["agent-governance-toolkit-integrations[openai-agents]>=4.1.0,<5.0"]
+langgraph = ["agent-governance-toolkit-integrations[langgraph]>=4.1.0,<5.0"]
+llamaindex = ["agent-governance-toolkit-integrations[llamaindex]>=4.1.0,<5.0"]
+haystack = ["agent-governance-toolkit-integrations[haystack]>=4.1.0,<5.0"]
+pydantic-ai = ["agent-governance-toolkit-integrations[pydantic-ai]>=4.1.0,<5.0"]
+adk = ["agent-governance-toolkit-integrations[adk]>=4.1.0,<5.0"]
 # Legacy extras (kept for backward compat)
-kernel = ["agent-governance-toolkit-core>=4.0.0,<5.0"]
-mesh = ["agent-governance-toolkit-core>=4.0.0,<5.0"]
-runtime = ["agent-governance-toolkit-core>=4.0.0,<5.0"]
-sre = ["agent-governance-toolkit-cli>=4.0.0,<5.0"]
+kernel = ["agent-governance-toolkit-core>=4.1.0,<5.0"]
+mesh = ["agent-governance-toolkit-core>=4.1.0,<5.0"]
+runtime = ["agent-governance-toolkit-core>=4.1.0,<5.0"]
+sre = ["agent-governance-toolkit-cli>=4.1.0,<5.0"]
 opa = []
 cedar = ["cedarpy>=4.0.0,<5.0"]
 full = [
-    "agent-governance-toolkit-core>=4.0.0,<5.0",
-    "agent-governance-toolkit-integrations>=4.0.0,<5.0",
-    "agent-governance-toolkit-cli>=4.0.0,<5.0",
-    "agent-governance-toolkit-protocols>=4.0.0,<5.0",
+    "agent-governance-toolkit-core>=4.1.0,<5.0",
+    "agent-governance-toolkit-integrations>=4.1.0,<5.0",
+    "agent-governance-toolkit-cli>=4.1.0,<5.0",
+    "agent-governance-toolkit-protocols>=4.1.0,<5.0",
 ]
 
 [project.urls]

agent-governance-python/agent-governance-toolkit-cli/pyproject.toml

@@ -35,7 +35,7 @@ classifiers = [
 
 # Merged from agent-sre + agent-sandbox + mcp-trust-server
 dependencies = [
-    "agent-governance-toolkit-core>=4.0.0,<5.0",
+    "agent-governance-toolkit-core>=4.1.0,<5.0",
     "pydantic>=2.4.0,<3.0",
     "pyyaml>=6.0,<7.0",
     "croniter>=2.0,<7.0",
@@ -46,7 +46,7 @@ dependencies = [
 [project.optional-dependencies]
 docker = ["docker>=7.1.0,<8.0"]
 hyperlight = ["hyperlight-sandbox>=0.4.0,<0.5"]
-policy = ["agent-governance-toolkit-core>=4.0.0,<5.0"]
+policy = ["agent-governance-toolkit-core>=4.1.0,<5.0"]
 mcp = ["mcp>=1.0.0,<2.0", "cryptography>=46.0.7,<49.0"]
 api = ["fastapi>=0.115.0,<1.0", "uvicorn>=0.27.0,<1.0"]
 otel = ["opentelemetry-exporter-otlp>=1.20,<2.0"]
@@ -63,7 +63,7 @@ wandb = ["wandb>=0.16,<1.0"]
 mlflow = ["mlflow>=2.10,<4.0"]
 agentops = ["agentops>=0.3,<1.0"]
 full = [
-    "agent-governance-toolkit-cli[docker,mcp,api,otel]",
+    "agent-governance-toolkit-cli[docker,mcp,api,otel,policy]",
 ]
 dev = [
     "pytest>=8.0,<10.0",

agent-governance-python/agent-governance-toolkit-integrations/pyproject.toml

@@ -8,7 +8,7 @@ version = "4.1.0"
 description = "Framework adapters for the Agent Governance Toolkit (LangChain, CrewAI, OpenAI Agents, and more)"
 readme = "README.md"
 license = {text = "MIT"}
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 authors = [
     {name = "Microsoft Corporation", email = "agentgovtoolkit@microsoft.com"},
 ]
@@ -26,7 +26,6 @@ classifiers = [
     "Topic :: Scientific/Engineering :: Artificial Intelligence",
     "Topic :: Software Development :: Libraries :: Python Modules",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Programming Language :: Python :: 3.13",
@@ -35,7 +34,7 @@ classifiers = [
 
 # No mandatory framework deps: each adapter is an optional extra.
 dependencies = [
-    "agent-governance-toolkit-core>=4.0.0,<5.0",
+    "agent-governance-toolkit-core>=4.1.0,<5.0",
 ]
 
 [project.optional-dependencies]
@@ -103,7 +102,7 @@ packages = ["_placeholder"]
 include = ["README.md", "LICENSE"]
 
 [tool.ruff]
-target-version = "py310"
+target-version = "py311"
 line-length = 100
 
 [tool.ruff.lint]

agent-governance-python/agent-governance-toolkit-protocols/pyproject.toml

@@ -8,7 +8,7 @@ version = "4.1.0"
 description = "Protocol implementations (MCP governance, trust protocol, A2A, MCP receipts) for the Agent Governance Toolkit"
 readme = "README.md"
 license = {text = "MIT"}
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 authors = [
     {name = "Microsoft Corporation", email = "agentgovtoolkit@microsoft.com"},
 ]
@@ -26,15 +26,14 @@ classifiers = [
     "Topic :: Security",
     "Topic :: Software Development :: Libraries",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Programming Language :: Python :: 3.13",
     "Typing :: Typed",
 ]
 
 dependencies = [
-    "agent-governance-toolkit-core>=4.0.0,<5.0",
+    "agent-governance-toolkit-core>=4.1.0,<5.0",
 ]
 
 [project.optional-dependencies]
@@ -65,7 +64,7 @@ packages = ["_placeholder"]
 include = ["README.md", "LICENSE"]
 
 [tool.ruff]
-target-version = "py310"
+target-version = "py311"
 line-length = 100
 
 [tool.ruff.lint]

agent-governance-python/agent-lightning/pyproject.toml

@@ -8,7 +8,7 @@ version = "4.1.0"
 description = "Public Preview — Agent-Lightning RL integration for the Agent Governance Toolkit: governed training with policy enforcement"
 readme = "README.md"
 license = {text = "MIT"}
-requires-python = ">=3.9"
+requires-python = ">=3.11"
 authors = [
     {name = "Microsoft Corporation", email = "agentgovtoolkit@microsoft.com"},
 ]
@@ -23,16 +23,14 @@ classifiers = [
     "Development Status :: 4 - Beta",
     "Intended Audience :: Developers",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.9",
-    "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Topic :: Scientific/Engineering :: Artificial Intelligence",
 ]
 dependencies = []
 
 [project.optional-dependencies]
-agent-os = ["agent-governance-toolkit-core>=4.0.0,<5.0"]
+agent-os = ["agent-governance-toolkit-core>=4.1.0,<5.0"]
 dev = ["pytest>=7.0,<10.0", "pytest-cov"]
 
 [project.urls]

agent-governance-python/agent-marketplace/pyproject.toml

@@ -8,7 +8,7 @@ version = "4.1.0"
 description = "Plugin marketplace for the Agent Governance Toolkit — discover, install, verify, and manage plugins"
 readme = "README.md"
 license = {text = "MIT"}
-requires-python = ">=3.9"
+requires-python = ">=3.11"
 authors = [
     { name = "Microsoft Corporation", email = "agentgovtoolkit@microsoft.com" }
 ]
@@ -21,8 +21,6 @@ classifiers = [
     "Intended Audience :: Developers",
     "License :: OSI Approved :: MIT License",
     "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.9",
-    "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Topic :: Software Development :: Libraries :: Application Frameworks",

agent-governance-python/agent-mcp-governance/pyproject.toml

@@ -7,7 +7,7 @@ name = "agent_mcp_governance"
 version = "4.1.0"
 description = "Deprecated. Previously published as agent-mcp-governance. Install agent-governance-toolkit-protocols instead."
 license = {text = "MIT"}
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 authors = [
     {name = "Microsoft Corporation", email = "agentgovtoolkit@microsoft.com"},
 ]

agent-governance-python/agent-os/README.md

@@ -1131,7 +1131,7 @@ export ANTHROPIC_API_KEY=sk-ant-...
 Prompt-based guardrails ask the LLM to self-police, which is probabilistic. Agent OS enforces governance at the middleware level using deterministic policy engines and POSIX-inspired access controls. It controls what agents *can* do (capability-based), not just what they *should not* do (filter-based).
 
 **How does Agent OS work with other frameworks?**
-Agent OS integrates with 14+ frameworks via adapters. Install the governance layer alongside your existing framework: use `langgraph-trust` for LangGraph, `agentmesh-openai-agents-trust` for OpenAI Agents, or the MCP server for any MCP-compatible client. Agent OS acts as a kernel layer underneath your agent framework.
+Agent OS integrates with 21 integration packages. Install the governance layer alongside your existing framework: use `langgraph-trust` for LangGraph, `agentmesh-openai-agents-trust` for OpenAI Agents, or the MCP server for any MCP-compatible client. Agent OS acts as a kernel layer underneath your agent framework.
 
 **What is the Agent Governance Ecosystem?**
 Agent OS is part of a suite of seven packages: Agent OS (policy engine), [AgentMesh](https://github.com/microsoft/agent-governance-toolkit) (trust infrastructure), [Agent Runtime](https://github.com/microsoft/agent-governance-toolkit) (execution supervisor), [Agent SRE](https://github.com/microsoft/agent-governance-toolkit) (reliability), [Agent Compliance](https://github.com/microsoft/agent-governance-toolkit) (regulatory compliance), [Agent Marketplace](https://github.com/microsoft/agent-governance-toolkit) (plugin lifecycle), and [Agent Lightning](https://github.com/microsoft/agent-governance-toolkit) (RL training governance). Together they provide 4,310+ tests across 17 modules.

agent-governance-python/agent-os/docs/comparison-with-alternatives.md

@@ -123,7 +123,7 @@ define flow
 #### Where Agent OS is better
 
 - **Tool call interception** — Agent OS intercepts the actual action, not just the LLM output that might trigger it
-- **Multi-framework support** — 12+ framework integrations vs NeMo's primary LangChain focus
+- **Multi-framework support** — 21 integration packages vs NeMo's primary LangChain focus
 - **Performance** — Sub-0.1ms overhead vs NeMo's 10–50ms+ per rail evaluation
 - **Multi-agent governance** — Per-agent policies, inter-agent trust, fleet management
 
@@ -224,7 +224,7 @@ result = guard(llm.complete, prompt="Summarize the report")
 | **Schema enforcement** | ❌ | ✅ Pydantic/JSON Schema |
 | **Auto-retry on failure** | ❌ | ✅ Corrective re-prompting |
 | **Tool call governance** | ✅ Core capability | ❌ Not applicable |
-| **Multi-framework** | ✅ 12+ frameworks | ⚠️ OpenAI, Anthropic, Cohere |
+| **Multi-framework** | ✅ 21 integrations | ⚠️ OpenAI, Anthropic, Cohere |
 | **PII detection** | ⚠️ Pattern-based | ✅ Multiple detector validators |
 | **Latency** | <0.1ms p99 | <50ms (async validators) |
 
@@ -238,7 +238,7 @@ result = guard(llm.complete, prompt="Summarize the report")
 #### Where Agent OS is better
 
 - **Action governance** — Agent OS governs what agents *do*, not just what they *say*
-- **Framework coverage** — 12+ framework integrations vs. 3 primary LLM providers
+- **Framework coverage** — 21 integration packages vs. 3 primary LLM providers
 - **Deterministic enforcement** — Policies block actions regardless of LLM output
 - **Multi-agent support** — Per-agent policies, inter-agent trust protocols
 
@@ -339,7 +339,7 @@ raise "Don't send email after reading website" if:
 | **Chain detection** | ⚠️ Per-call policies | ✅ Multi-step chain rules |
 | **Built-in detectors** | Pattern matching | PII, secrets, copyright, prompt injection |
 | **Trace visualization** | ⚠️ Audit logs | ✅ Invariant Explorer |
-| **Framework support** | 12+ integrations | 2–3 (LLM/MCP proxy) |
+| **Framework support** | 21 integrations | 2–3 (LLM/MCP proxy) |
 | **Deployment** | Embedded library | Library or proxy |
 | **Multi-agent** | ✅ | ❌ Single-agent |
 | **Human-in-the-loop** | ✅ | ❌ |
@@ -353,7 +353,7 @@ raise "Don't send email after reading website" if:
 
 #### Where Agent OS is better
 
-- **Framework coverage** — 12+ framework integrations vs. proxy-only approach
+- **Framework coverage** — 21 integration packages vs. proxy-only approach
 - **Multi-agent governance** — Per-agent policies, inter-agent trust, fleet management
 - **Human-in-the-loop** — Native human approval workflows
 - **Ecosystem** — Part of a broader governance stack (AgentMesh, Agent SRE, Agent Runtime)
@@ -400,7 +400,7 @@ Cordum is a **standalone Agent Control Plane** built in Go. It provides a Before
 #### Where Agent OS is better
 
 - **Lightweight integration** — `pip install` vs. Docker Compose infrastructure
-- **Framework diversity** — 12+ framework adapters vs. CAP protocol adoption
+- **Framework diversity** — 21 integration packages vs. CAP protocol adoption
 - **Action-level granularity** — Per-tool-call policies vs. job-level gating
 - **Open license** — MIT vs. BUSL-1.1 (commercial restrictions)
 - **Ecosystem maturity** — Merged into Dify (65K★), LlamaIndex (47K★), Agent-Lightning (15K★)
@@ -490,7 +490,7 @@ TrinityGuard is a **multi-agent safety testing and monitoring framework** that c
 #### Where Agent OS is better
 
 - **Runtime enforcement** — Agent OS *blocks* dangerous actions; TrinityGuard primarily *detects* them
-- **Framework coverage** — 12+ integrations vs. AG2/AutoGen focus
+- **Framework coverage** — 21 integrations vs. AG2/AutoGen focus
 - **Performance** — <0.1ms deterministic checks vs. LLM inference latency
 - **Production readiness** — 1,680+ tests, benchmark suite, enterprise adoption
 

agent-governance-python/agent-os/examples/carbon-auditor/pyproject.toml

@@ -4,7 +4,7 @@ version = "4.1.0"
 description = "Autonomous auditing system for the Voluntary Carbon Market (VCM)"
 license = {text = "MIT"}
 readme = "README.md"
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 dependencies = [
     "numpy>=1.20.0,<2.0",
     "pydantic>=2.4.0,<3.0",

agent-governance-python/agent-os/examples/defi-sentinel/pyproject.toml

@@ -7,7 +7,7 @@ name = "defi_sentinel_demo"
 version = "4.1.0"
 description = "DeFi Risk Sentinel demo using Agent OS"
 license = {text = "MIT"}
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 dependencies = [
     "pydantic>=2.4.0,<3.0",
 ]

agent-governance-python/agent-os/examples/grid-balancing/pyproject.toml

@@ -7,7 +7,7 @@ name = "grid_balancing_demo"
 version = "4.1.0"
 description = "Autonomous energy trading demo using Agent OS"
 license = {text = "MIT"}
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 dependencies = [
     "pydantic>=2.4.0,<3.0",
     "numpy>=1.20.0,<2.0",

agent-governance-python/agent-os/examples/pharma-compliance/pyproject.toml

@@ -7,7 +7,7 @@ name = "pharma_compliance_demo"
 version = "4.1.0"
 description = "Pharma Compliance demo using Agent OS"
 license = {text = "MIT"}
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 dependencies = [
     "pydantic>=2.4.0,<3.0",
 ]