Agent Framework

What it does

When your agent uses it

• : creating AgentDefinition
• Adding function tools with AIFunctionFactory
• Connecting MCP servers
• Managing agent conversations and session state

Example prompts

• “/agent-framework”

Workflow steps

7 steps, taken from the step headings in SKILL.md.

1. 1Creating a New Agent (Backend)
2. 2Adding a Builtin Function Tool
3. 3Wiring a New LLM Provider
4. 4Adding Knowledge Items
5. 5Connecting an MCP Server
6. 6Running a Conversation Turn
7. 7Adding a New API Endpoint

What it can do on your machine

Read from SKILL.md and the folder at commit a69d04b. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • ollama

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Agent Framework loads about 1.9k tokens when it runs, and up to ~2.8k if it reads all its reference files. Until then it costs about 123 tokens; SKILL.md has 503 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~123

When it runs · the whole SKILL.md, loaded when a task matches

~1.9k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~2.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from jihadkhawaja/Egroo at commit a69d04b, republished under its Apache-2.0 licence (© jihadkhawaja). 503 words, ~1,915 tokens.

name

agent-framework

description

Build, extend, and debug AI agents in Egroo using the Microsoft Agent Framework (C# .NET). Use for: creating AgentDefinition, adding function tools with AIFunctionFactory, connecting MCP servers, managing agent conversations and session state, wiring new LLM providers (OpenAI, AzureOpenAI, Anthropic, Ollama), adding knowledge items, implementing streaming chat. Also covers AgentRuntimeService internals, BuiltinTools, and McpClientService patterns already in the codebase.

argument-hint

Describe what you want to build or fix (e.g. "add a weather tool", "new Ollama agent", "debug streaming")

Client (Blazor)
  └─ AgentService (HTTP)
       └─ POST /api/v1/Agent/conversations/{id}/send-message
            └─ AgentRuntimeService.ChatAsync / ChatStreamAsync
                 ├─ IAgentRepository  → loads AgentDefinition, knowledge, tools, history
                 ├─ EncryptionService → decrypts agent API key
                 ├─ AIAgent (Microsoft.Agents.AI)
                 │    └─ provider: OpenAI | AzureOpenAI | Anthropic | Ollama
                 ├─ BuiltinTools     → pre-built AIFunctions
                 └─ McpClientService → dynamic MCP tool proxies

public static AIFunction MyNewTool { get; } = AIFunctionFactory.Create(
    ([Description("The input")] string input) =>
    {
        // ... logic ...
        return Task.FromResult("result");
    },
    name: "my_new_tool",
    description: "What this tool does.");

var tools = new List<AIFunction> { BuiltinTools.GetCurrentDatetime, BuiltinTools.MyNewTool };

LlmProvider.Ollama => new OllamaApiClient(new Uri(definition.Endpoint!))
    .AsChatClient(definition.Model)
    .AsAIAgent(instructions: systemPrompt, tools: tools),

var knowledgeItems = await _agentRepo.GetAgentKnowledgeAsync(agentId);
var systemPrompt = definition.Instructions + "\n\n" +
    string.Join("\n", knowledgeItems.Where(k => k.IsEnabled).Select(k => k.Content));

// Non-streaming
var response = await _agentRuntimeService.ChatAsync(userId, conversationId, "Hello");

// Streaming (IAsyncEnumerable<string>)
await foreach (var token in _agentRuntimeService.ChatStreamAsync(userId, conversationId, "Hello"))
{
    // push token to client via SignalR or SSE
}

group.MapPost("/{agentId}/my-feature", async (IAgentRepository repo, Guid agentId, MyRequest req, HttpContext ctx) =>
{
    // use repo, return Results.Ok/NotFound
})
.RequireAuthorization();

© jihadkhawaja, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file (references) in .github/skills/agent-framework of jihadkhawaja/Egroo.

• SKILL.md
• references/ms-agent-framework.md

Open the folder on GitHubat commit a69d04b