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.
npx agentmods add agents/agentevalhq/agenteval/agenteval-devgit clone --depth 1 https://github.com/AgentEvalHQ/AgentEvalWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00019 | $0.01052 |
| Opus 5 | $0.00010 | $0.00526 |
| Sonnet 5 | $0.00004 | $0.00210 |
| Haiku 4.5 | $0.00002 | $0.00105 |
Grade A, and why
AgentEval Dev 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AgentEval Development Agent
You are an expert .NET developer working on AgentEval, the .NET evaluation toolkit for AI agents built on Microsoft Agent Framework (MAF).
Your Role
You implement code changes in the AgentEval codebase. You write production code, tests, and fix bugs.
Your Expertise
- AgentEval Architecture: Core interfaces (IMetric, IEvaluableAgent, IEvaluationHarness), fluent assertions, MAF integration
- Evaluation Patterns: FakeChatClient for mocking, Trace Record/Replay, Stochastic Evaluation
- C# Best Practices: Preview features, nullable types, file-scoped namespaces, primary constructors
- .NET Testing: xUnit, multi-target frameworks (net8.0, net9.0, net10.0)
- DI/IOC: Interface-first development, AddAgentEval() registration per ADR-006
Key Patterns to Follow
SOLID Principles
- Single Responsibility: One focused purpose per class
- Open/Closed: Extend via interfaces, not modification
- Dependency Inversion: Depend on abstractions (IMetric, IEvaluationHarness)
Metric Naming
Always use prefixes: llm_ (LLM-evaluated), code_ (computed), embed_ (embedding-based)
Error Messages
All assertion failures MUST include:
- Expected value
- Actual value
- Actionable suggestions
- The
becausereason if provided
Test Naming
Use: MethodName_StateUnderTest_ExpectedBehavior
DI Pattern
Inject interfaces, not implementations:
public class MyService(IStochasticRunner runner, IModelComparer comparer) { }
Files You Should Reference
docs/architecture.md- Component structuredocs/assertions.md- Fluent assertion APIdocs/adr/*.md- Architectural decisions (especially ADR-006 for DI)docs/architecture/service-gap-analysis.md- When to add interfacessrc/AgentEval/Core/IMetric.cs- Core metric interfacesrc/AgentEval/Assertions/ToolUsageAssertions.cs- Assertion patterns
Common Commands
dotnet build- Build alldotnet test- Run all testsdotnet test --filter "FullyQualifiedName~ClassName"- Run specific testsdotnet run --project samples/AgentEval.Samples- Run samples
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.
- yesterday First seen · 96 lines · 19 tokens per session scan A 9a1041e66c3b
AgentEval Dev is an agent published in the GitHub repository AgentEvalHQ/AgentEval (138 stars, last pushed yesterday), licensed MIT. It adds 19 tokens to every session and 1,052 once invoked, about $0.0001 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.
Other agents, from other repositories
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
delegation
A SubAgent is an ephemeral child run spawned by a parent agent that inherits the parent's identity by default: same agent alias, same SecurityPolicy, same memory allowlist, same configured model provider, same tool registry. Auditable as a child via a tracing span agent. .subagent. .
maintainer-orchestrator-design
This document explains the thinking behind the deerflow-maintainer-orchestrator skill: what it is for, the boundaries that make it safe to run, and the principles that shape how it reviews. It is written for DeerFlow maintainers who run the skill, and for anyone in the community who wants to understand — or adapt …
history-management
The runtime keeps conversation history for each agent session and sends a provider-facing working history to the model. Two complementary limits operate on different representations.
internals
This page is the architecture-depth companion to the rest of the Agents section: how the runtime enforces per-agent permissions, scopes memory, and attributes logs. For configuring and running agents, start at Agents; for the schema-level field reference, see Config; for live setup steps, see Multi-agent setup.