> For the complete documentation index, see [llms.txt](https://docs.zerve.ai/zerve-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.zerve.ai/zerve-docs/getting-started/key-concepts.md). # Key concepts The terms used throughout the Zerve docs. These are the terms used throughout Zerve and these docs. ## Superagents | Term | Meaning | | ------------------ | ------------------------------------------------------------------------------------------------------------------------------- | | **Superagent** | A long-running agent that owns a mission and manages a team of agents working toward it. | | **Mission** | The starting prompt a superagent is created from, written in plain language. | | **Objective** | A goal the superagent is working toward. Each objective has a stage, a status and an autonomy level. | | **KPI** | A measurable success criterion, such as a metric with a target. KPIs can be tied to an objective or stand on their own. | | **Autonomy level** | How freely a superagent can make decisions on an objective: Ask me first, Balanced or Full autonomy. | | **Coordinator** | A member of a superagent's team that manages one branch of the work and delegates to its own team members. | | **Worker** | A member of a superagent's team that does the hands-on work in its own notebook. Every worker is an Evolve Mode agent. | | **Subagent** | A one-off specialist a superagent runs for a self-contained job, such as research or an independent check. | | **Batch** | A group of runs a superagent carries out in parallel, such as a tournament between competing approaches. | | **Recommendation** | A decision or input the superagent needs from you. Shown in the app as a decision waiting. | | **Backlog** | The superagent's list of open work items. | | **Artifact** | A versioned file a superagent's team produces, such as a dataset, model or scored result, that other team members can build on. | | **Memory** | What a superagent has learned and decided, kept across conversations. | | **Channel** | A way to reach a superagent from outside Zerve, such as Slack or Telegram. | ## Agents in the Workspace | Term | Meaning | | --------------- | ----------------------------------------------------------------------------------------------------------------------------- | | **Thread** | A standalone conversation with an agent. | | **Agent mode** | How much the agent does on its own: Ask, Plan, Build or Evolve. | | **Evolve Mode** | The most capable agent mode. The agent plans, delegates to specialists in parallel, verifies the results and keeps learnings. | | **Specialist** | An agent Evolve Mode delegates to, such as a coding, research or review agent. Also called a subagent. | | **Learnings** | An agent's persistent notes on a notebook: its objectives, assumptions, approach and research topics. | | **Skill** | A reusable set of instructions and best practices an agent can follow, written as a SKILL.md file. | | **Sources** | The research references an agent saves and cites. | ## Notebooks and compute | Term | Meaning | | ---------------- | ----------------------------------------------------------------- | | **Notebook** | Where code, data and results live, organized as connected blocks. | | **Block** | A single step in a notebook, such as a piece of Python, R or SQL. | | **Block run** | A single execution of a block on compute. | | **Compute type** | Where a block runs, such as Lambda, Fargate or a GPU. | | **Fleet** | Zerve's way of running a block many times in parallel. | | **Environment** | The packages and settings a block runs with. | ## Assets and administration | Term | Meaning | | ---------------- | ------------------------------------------------------------------------------------------------ | | **Asset** | A reusable item shared across notebooks: a data connection, a constant or secret, or a function. | | **Organization** | A group of people who share work, assets, settings and a plan. | | **Credits** | The unit Zerve uses to measure usage of agents and compute. | --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.zerve.ai/zerve-docs/getting-started/key-concepts.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.