DevSocial Articles · AI Agents articles

Learn how to test AI agents for indirect prompt injection in tool results, verify tool authorization and side effects, and build adversarial xUnit tests in C#.

Aarav PatelOct 232,2701,000

Learn how to restrict AI coding agents to approved terminal operations in C# using allowlists, a command broker, and safe process execution.

Aarav PatelOct 232,4541,000

Explains AI software factories: coordinated AI agents that plan, build, test, review, deploy, and monitor software with human oversight and automated validation.

THE TISAOct 234,1761,000

Learn how to track token usage and model costs across delegated AI agents in C#, using per-request usage records and configurable pricing.

Aarav PatelOct 232,7211,000

Learn how to design auditable AI agent workflows in C# and .NET using structured audit events, authorization records, correlation IDs, and durable audit storage.

Ananya DesaiOct 234,2111,000

Learn how to run multiple AI coding agents safely using Git worktrees, clear task boundaries, shared contracts, and controlled code integration.

Aarav PatelOct 234,2311,000

Learn how to test AI agent answers with evaluation datasets, automated checks, semantic criteria, tool-use validation, and safe failure handling.

Aarav PatelOct 222,6001,000

Enterprise artificial intelligence is progressing from isolated conversational systems toward architectures in which specialized agents can maintain context, apply professional capabilities, use tools, collaborate with other agents, and participate directly in organizational work. This development changes the central architectural question. The question is no longer simply: How capable is the underlying language model? It increasingly becomes: How should model intelligence be structured, specialized, governed, organized, and operationalized inside an enterprise? Gate2ASI AI’s AgentFactory, originally developed as AlpineGate AI’s AgentFactory, addresses this question through two related concepts developed by John Godel: GSCP-15 — Gödel’s Scaffolded Cognitive Prompting and PT-SLM — Private Tailored Small Language Model.

John GodelOct 222,0571,000

Explores how AgentFactory applies metacognitive monitoring, evidence, validation, and governance to build accountable autonomous AI agents.

John GodelOct 212,1741,000

Learn how a BFF uses signed continuation tokens to bind Microsoft Foundry Responses conversations to authenticated users and prevent unauthorized reuse of response IDs.

Cristopher CoronadoOct 203,0811,000

Learn how Microsoft Foundry IQ, Fabric IQ, Work IQ, and Memory Stores support agentic enterprise retrieval, grounded answers, and long-term user memory.

NithinOct 192,0371,000

Learn how to centralize reusable agent tools with Microsoft Foundry Toolboxes, MCP endpoints, and Tool Search to reduce prompt tokens and simplify tool management.

NithinOct 192,5181,000

Learn how Microsoft Foundry Agent Service supports declarative prompt agents, real-time voice agents, and hosted Python or .NET agents deployed from ZIP files or containers.

NithinOct 193,9031,000

Learn how Microsoft Foundry supports production AI agents with unified project endpoints, the Responses API, managed tools, stateful conversations, and the Azure.AI.Projects SDK.

NithinOct 194,6941,000

Learn why AI agents exceed task boundaries and how to constrain them with task contracts, least-privilege access, authorization controls, and prompt-injection defenses.

Mahesh ChandOct 173,6171,000

Learn how REA MCP Server connects AI coding agents to reverse-engineering tools to inspect binaries, Electron apps, and .NET assemblies without source code.

Ananya DesaiOct 152,9541,000

Learn how to build an enterprise RFP assistant with Microsoft Copilot Studio, Azure AI, Power Automate, and Python to analyze documents, coordinate stakeholders, and draft proposals.

Dhanasekaran GOct 142,7791,000

AI agents are moving beyond short conversations. Developers are building agents that inspect codebases, execute tools, run tests, investigate failures, and complete workflows that may take hours to finish. However, running an agent in production introduces a different set of engineering challenges.

Nidhi SharmaOct 133,2091,000

Learn how Google's Gemini agent coordinates work across Gmail, Drive, Docs, Calendar, and other Workspace apps, including permissions, governance, and project-status reporting workflows.

Vivek KumarOct 122,3211,000

Learn how to build long-term AI agent memory with AlloyDB for durable, searchable records and Memorystore for Valkey for low-latency state and caching.

Niharika GuptaOct 93,3581,000

Explore how Google's Gemini agent works across Workspace apps, coordinates multi-step tasks and sub-agents, and raises governance considerations for enterprise teams.

Aarav PatelOct 93,5791,000

Build a controlled .NET AI agent harness with Microsoft.Extensions.AI, including tool calling, conversation history, cancellation, error handling, and bounded execution.

Riya PatelOct 92,1181,000

Explores shared memory and reusable skills for AI agents, including memory types, ownership, retrieval, expiration, and a C# model for storing approved agent context.

Ananya DesaiOct 93,5081,000

Learn how to use Microsoft Foundry's agent optimizer to evaluate agent configurations, generate improvements, compare scores, and apply optimized candidates.

NithinOct 83,8351,000

Learn how Microsoft Foundry Toolboxes and tool search reduce agent token usage, improve tool selection, and support MCP tools in Python and C#.

NithinOct 83,2671,000

Build a Microsoft Foundry agent that uses MCP tools, Cosmos DB session state, approval gates, and OpenTelemetry tracing for enterprise-ready workflows.

NithinOct 82,4561,000

Learn how Azure Functions agent bindings let Python apps run event-driven AI agent workflows with tool calling, security boundaries, and business-rule enforcement.

Nidhi SharmaOct 83,4611,000

Learn how Microsoft Execution Containers on Windows 11 isolate AI coding agents, restrict host access, and reduce risks from untrusted code and prompt injection.

Bhavya SinghOct 84,8921,000

Building an AI agent is only the beginning. The harder question is: how do you understand what your agent is actually doing in production, and find problems you didn't know existed? Microsoft has introduced Insights in Foundry as a public preview capability. It analyzes production agent traces, iden

Monika MundraOct 82,4031,000