> ## Documentation Index > Fetch the complete documentation index at: https://docs.corepillar.ai/llms.txt > Use this file to discover all available pages before exploring further. # Introduction > AgentHub is CorePillar's platform for tracing, costing, and evaluating AI agents, served at pillarforge.corepillar.ai. AgentHub is where your team watches AI agents work. Your agent sends a record of each run to AgentHub, AgentHub stores every step, prices it, and lets you grade it against evaluators you choose. You can also build an agent from a template, generate test scenarios for it, and run those scenarios against real models to see what each one costs before you ship. AgentHub runs at [https://pillarforge.corepillar.ai](https://pillarforge.corepillar.ai). AgentHub never runs your agent's code. Your system emits traces and AgentHub receives them. The only model calls AgentHub makes on its own are the ones you ask for: LLM-judge evaluations, Cost Calculator runs, and synthetic data generation. ## Who it is for AgentHub is for **admins** and **AI Builders** in your organization. People with the **User** role (the Viewer database role) can sign in, but every AgentHub page shows "Your account has ModelHub access only" and a **Go to ModelHub** button. See [Roles and permissions](/agenthub/settings/roles-and-permissions). ## The core loop ```mermaid theme={null} flowchart LR A[Instrument your agent] --> B[Send traces] B --> C[Inspect in Traces] C --> D[Evaluate against a collection] D --> E[Report, compare models, export] E --> A ``` 1. **Instrument.** Create a tracing key, then have your agent POST its spans to AgentHub's ingest endpoint as JSON. There is no published SDK yet; ingestion is a REST endpoint. See the [tracing quickstart](/agenthub/developers/tracing-quickstart). 2. **Trace.** Each run shows up under **Traces** with its spans, tokens, latency, cost, and status. 3. **Evaluate.** Score a trace against an evaluator collection. Offline evaluators are deterministic and free; LLM-judge evaluators call a model you have configured. 4. **Optimize and export.** Use the Cost Calculator and Simulations to compare models on the same scenarios, write up results in Reports, and preview reinforcement-learning exports. ## What is in the sidebar The AgentHub sidebar, top to bottom: | Item | What it does | | - | - | | **Overview** | Project home: headline numbers, recent runs, and an integration guide with your project's masked tracing key. | | **Model Manager** | Store provider API keys in the encrypted vault and choose which models this project can use. | | **AI Usage** | Organization-wide spend and usage dashboard. On AgentHub it sits where the Playground would be. Admins see the whole organization; AI Builders see only their own usage. | | **Cost Calculator** | Describe an agent, generate scenarios, and run them against real models to estimate cost, latency, and quality. | | **Evaluation** | Evaluators, collections, and evaluation runs. | | **Simulations** | Past Cost Calculator runs with grader results and pass/fail breakdowns. | | **Traces** | Every trace your agents sent, with a detail view per trace. | | **Reports** | Readable write-ups generated from a Cost Calculator run, a trace evaluation, or a trace. | **Settings** sits at the bottom of the sidebar and opens as a window over the page you are on. Only admins and super admins see it. See [Settings overview](/agenthub/settings/overview). The model comparison Playground exists only on CorePillar's demo deployment. On AgentHub, that slot is the **AI Usage** dashboard. ## Module guides The project home page and integration guide. Provider keys, the Platform Demo Key, and project models. Spend by person and project, token split, and privacy events. Build an agent, generate scenarios, and run them on real models. Review past Cost Calculator runs and their grades. List, filter, and inspect traces span by span. Evaluators, collections, and runs. Turn source documents into an agent spec, records, and scenarios. Shareable write-ups of runs, evaluations, and traces. Label trajectories and preview a JSONL export. Synthetic data and RL export are not in the sidebar. Their guides explain how to reach them and how complete they are today. ## AgentHub and ModelHub CorePillar ships two products from one platform. They share the same accounts, organizations, projects, and model configurations. | | ModelHub | AgentHub | | - | - | - | | URL | [modelhub.corepillar.ai](https://modelhub.corepillar.ai) | [pillarforge.corepillar.ai](https://pillarforge.corepillar.ai) | | Audience | Every employee | Admins and AI Builders | | What it offers | Governed AI chat and personal usage | Tracing, evaluation, cost analysis, and administration | | Who configures models | Nobody on ModelHub; models come from AgentHub | Admins and AI Builders, in Model Manager | The models your employees chat with on ModelHub are the ones configured in AgentHub's Model Manager, and their chat spend appears in AgentHub's AI Usage dashboard. Membership belongs to the organization, not to a site, so someone invited to your organization can sign in to both. What they can do on each is set by their role. See the [ModelHub introduction](/modelhub/introduction). ## Next steps From sign-in to your first evaluated trace. Organizations, projects, roles, keys, traces, and evaluators. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.