Documentation
Guides and reference for iAgents Hub
Key terms
An AI assistant configured for a specific function. It can answer questions, draft content, analyze data, or automate tasks — always within the boundaries you define.
The AI engine that processes text and generates responses. It is the underlying technology — Claude, GPT-4o, Gemini… — that the agent uses to operate.
The text that defines the agent's role, tone, and behavioral rules. It is what turns a generic assistant into one specialized for your use case.
Configured access to a specific LLM. Assigning a connection to an agent determines which AI engine it will use to respond.
The company that offers the LLM: Anthropic (Claude), OpenAI (GPT), Google (Gemini), and others. You supply your own API key and pay the provider directly based on usage.
An additional capability activated on an agent: following a style guide, restricting topics, enforcing a specific output format… Skills are reusable across agents.
Allows the agent to remember information from previous conversations with the same user. Useful for maintaining continuity without repeating context each session.
Your own documents, web pages, or texts that the agent can consult when responding. The content is not in the instructions — it is retrieved only when relevant.
The unit providers use to measure and charge LLM usage. Roughly equivalent to one word. The more text exchanged per conversation, the more tokens are consumed and the higher the cost.
A parameter that controls how creative the model is. Low value: consistent, predictable responses. High value: more varied and imaginative responses. Configured per agent.
The maximum amount of text an LLM can read and consider at once: instructions, message history, and included documents. Exceeding it causes the model to "forget" the beginning of the conversation.
When the model generates incorrect or fabricated information with complete confidence. This is not an occasional glitch — it is an inherent limitation of all LLMs. Responses must be reviewed in high-stakes contexts.
Capabilities the agent can invoke to act in the real world: searching the web, querying a database, calling an external API, or running code. They are what distinguishes an agent that only answers from one that also acts.
Retrieval-Augmented Generation. A technique that lets the agent search through your own documents before responding, rather than relying solely on what the model learned during training. It is the technology behind the knowledge base.
The process of retraining an LLM on your own data to specialize it for a specific domain. This is different from configuring an agent: fine-tuning modifies the model itself, which involves significant cost, time, and technical resources.
Models capable of processing not just text, but also images, audio, or video. Several LLMs available in iAgents Hub are already multimodal (GPT-4o, Claude 3), enabling use cases such as analysing screenshots or describing scanned documents.
Getting started
Full, permanent access. Your username is public; your sign-in email stays private unless you choose to show it on your profile.
Lets you explore the platform without signing up. Session data (history, settings) is not kept after logging out.
iAgents Hub is a self-hosted AI agent platform, accessible from any device — the interface is responsive and adapts to both mobile and desktop. Follow these three steps to get started:
Go to Profile → Providers. Enter your API key for Anthropic, OpenAI, Google, or another supported provider. Click «Test» to verify the connection and «Sync» to import available models.
Go to Agents and click «New». From there you can create an agent from scratch, load one from a file (.md or .json), or pick one from the catalog. Choose the agent type, the model, and write the system instructions. You can add skills, memory, and a knowledge base.
In the agent list, click on an agent and select «New chat». Your conversation history is saved automatically.
Agents
Every AI platform (Claude, ChatGPT, Copilot…) has its own agent format. iAgents Hub abstracts that difference: you configure an agent once — instructions, skills, memory, knowledge base — and can test it in the browser with any LLM before exporting it to the target platform.
Connect any provider and switch the agent's model with one click. Compare responses between Claude, GPT-4o, or others without reconfiguring anything.
The system prompt, skills, and context are defined once in iAgents Hub. Export the agent to the native format of each target platform — Claude Code, OpenAI Assistants, or GitHub Copilot — or import existing agents with a content preview before confirming.
Name, system instructions (system prompt), model and active connection, enabled skills, persistent memory, knowledge base, and scheduled routines.
When memory is enabled, the agent remembers context from previous conversations. You can view and edit memories from the Memory section.
Automate tasks the agent should run on a regular basis. Define a name, an instruction, and a trigger — manual, webhook, or a cron expression for scheduled execution — without needing an active conversation.
Connections
A connection is access to a specific LLM from a provider, ready to be assigned to an agent.
A provider (in Profile → Providers) stores your API key and gives access to its model catalog. Syncing it automatically creates a connection for each available LLM. Assign whichever connection you prefer to each agent.
LLMs charge by tokens: text units of roughly 4 characters each. Each connection accumulates the total input and output tokens consumed. You can view this in the Dashboard and in the Connections list.
Skills
Skills are reusable instruction blocks that are injected into the agent's system prompt when active. Their category comes from a closed catalog; free-form tags are not supported.
They include read-only official skills and skills published by registered users. Everyone, including guests, can browse them.
You can create skills in the Skills section and choose whether they are private or public. In guest mode they are private, temporary, and kept only in session memory.
When editing an agent, the Skills tab shows the list of available skills. Enable the ones you want to include in the agent.
Groups & resource sharing
Groups let you share private resources — agents, skills, connections, and knowledge — with other users in a controlled way. Only the owner decides what to share and with whom.
Go to Profile → Groups and click «Create group». You become the owner automatically. Only registered accounts can create groups; guests do not have access to this feature.
From the group view, click «Invite» and enter the user's public username. They will receive a notification to accept or decline. Group owners and admins can invite and remove members.
There are two ways to share a resource with a group: use the share button on the resource card and select the group; or drag the card directly onto the group name in the left panel — when you drop it, the resource is shared instantly. For agents, the system also shares in cascade their associated private skills and knowledge bases.
Only three parties can remove a resource from a group: the original resource owner, the group owner, and platform administrators. Other group members can use the resource but cannot revoke sharing.
Resources others have shared with you appear in the corresponding sections with a «Shared» badge. You can use them and open them from the groups panel, but cannot edit or export them — ownership stays with the original creator.
Users without an account (guest mode) cannot create or join groups. Register an account to access collaboration features.
Memory & Knowledge
iAgents Hub has two ways to give an agent persistent context: conversational memory and a knowledge base.
Memory stores text snippets that the agent remembers between conversations. Useful for user preferences, context data, or notes. Enabled per agent.
The knowledge base stores documents (PDF, text, URL). The agent can search them when responding. Useful for technical documentation, FAQs, or any reference content.
Agent best practices
Applying these practices reduces token consumption and improves response quality.
Write short, focused system prompts. Avoid repeating instructions already implicit in the model. Every unnecessary line consumes tokens on every message.
Use lightweight models (Haiku, GPT-4o mini) for simple tasks like classification or short summaries. Reserve large models (Opus, GPT-4o) for complex reasoning or critical code generation.
Skills are only injected when active. If you have instructions that don't apply to every conversation, turn them into a skill and activate it per agent rather than always including them in the system prompt.
Don't paste entire documents into the system prompt. Load content into the agent's knowledge base: only the relevant part is retrieved for each message, drastically reducing input token usage.
Memory injects context from previous conversations on every turn. If the agent performs independent tasks (text generation, translation), disable it to avoid accumulating unnecessary tokens.
Use low temperature (0.0–0.3) for precision tasks: code, data extraction, classification. Use higher values (0.7–1.0) for creative writing or brainstorming, where variability adds value.