AgentEval dependency-injection.instructions.md

A set of coding rules for fluent assertions, which are readable checks written as chains such as “should have called this tool.” It defines where new checks belong and how they should report failures.

In plain words
What is it for?
Use it when adding checks for tool usage, performance, or responses in AgentEval tests.
Why use it?
It keeps assertion methods consistent, readable, and useful when a test fails. Clear failure messages make incorrect agent behavior easier to diagnose.

Instructions file for GitHub Copilot

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/agentevalhq/agenteval/dependency-injection
Clone the repo
git clone --depth 1 https://github.com/AgentEvalHQ/AgentEval

Made for: GitHub Copilot.

Per session 850 This file is loaded in full into every session.
When invoked 850 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00850 $0.00850
Opus 5 $0.00425 $0.00425
Sonnet 5 $0.00170 $0.00170
Haiku 4.5 $0.00085 $0.00085

Measured 2d ago against content hash 0f889f36cb04, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

AgentEval dependency-injection.instructions.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.github/instructions/dependency-injection.instructions.md · 144 lines

What it actually says

---
applyTo: "src/AgentEval/DependencyInjection/**/*.cs"
description: Guidelines for dependency injection and service registration
---

# Dependency Injection Guidelines

## Core Principle: Interface-First Development

All core services must depend on abstractions, not concretions. See ADR-006 for full details.

## Service Registration Pattern

```csharp
public static class AgentEvalServiceCollectionExtensions
{
    public static IServiceCollection AddAgentEval(
        this IServiceCollection services,
        Action<AgentEvalServiceOptions>? configure = null)
    {
        var options = new AgentEvalServiceOptions();
        configure?.Invoke(options);
        
        // Register singleton services (stateless)
        services.AddSingleton<IStatisticsCalculator, DefaultStatisticsCalculator>();
        services.AddSingleton<IToolUsageExtractor, DefaultToolUsageExtractor>();
        
        // Register scoped services (stateful per-operation)
        services.AddScoped<IStochasticRunner, StochasticRunner>();
        services.AddScoped<IModelComparer>(sp => 
            new ModelComparer(sp.GetRequiredService<IStochasticRunner>()));
        
        return services;
    }
}

Service Lifetime Guidelines

Singleton (stateless)

Use for services that:

  • Have no mutable state
  • Are thread-safe
  • Can be shared across all requests

Examples: IStatisticsCalculator, IToolUsageExtractor

Scoped (stateful per-operation)

Use for services that:

  • Maintain state during a single operation
  • Need fresh instance per test run
  • Have dependencies on other scoped services

Examples: IStochasticRunner, IModelComparer

Transient (new each time)

Use sparingly for:

  • Lightweight, disposable objects
  • Objects with very short lifecycles

When NOT to Register as Services

Per service-gap-analysis.md, these should NOT be in DI:

  • Builders: AgentEvalBuilder (fluent API - direct instantiation)
  • Configuration POCOs: StochasticOptions, ModelComparisonOptions
  • Test-time tools: PerformanceBenchmark, SnapshotComparer

Adding a New Service

  1. Define interface in Core/ or domain folder:

    public interface IMyService
    {
        Task<Result> DoWorkAsync(Input input);
    }
    
  2. Implement interface:

    public class MyService : IMyService
    {
        private readonly IDependency _dependency;
        
        public MyService(IDependency dependency)
        {
            _dependency = dependency;
        }
        
        public async Task<Result> DoWorkAsync(Input input) { ... }
    }
    
  3. Register in DI:

    services.AddScoped<IMyService, MyService>();
    
  4. Inject via constructor (never resolve manually):

    public class Consumer(IMyService myService) { }
    

SOLID Principles in DI

Dependency Inversion (D in SOLID)

// ❌ BAD: Depending on concretion
public class ModelComparer
{
    private readonly StochasticRunner _runner; // concrete!
}

// ✅ GOOD: Depending on abstraction
public class ModelComparer : IModelComparer
{
    private readonly IStochasticRunner _runner; // interface!
}

Interface Segregation (I in SOLID)

// ✅ GOOD: Separate interfaces for distinct capabilities
public interface ITestableAgent { Task<string> ExecuteAsync(string prompt); }
public interface IStreamableAgent : ITestableAgent { IAsyncEnumerable<string> StreamAsync(string prompt); }

Testing with DI

For unit tests, inject mocks directly:

var mockRunner = new Mock<IStochasticRunner>();
var comparer = new ModelComparer(mockRunner.Object);

For integration tests, use the full DI container:

var services = new ServiceCollection();
services.AddAgentEval();
var provider = services.BuildServiceProvider();
var runner = provider.GetRequiredService<IStochasticRunner>();
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 144 lines · 850 tokens per session scan A 0f889f36cb04

Subscribe to this mod's changes

AgentEval dependency-injection.instructions.md is an instructions file published in the GitHub repository AgentEvalHQ/AgentEval (138 stars, last pushed 2d ago), licensed MIT. It adds 850 tokens to every session, about $0.0042 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.