4 min read·Updated 2026-04-10

Glossary

Canonical one-line definitions for the vocabulary used across the docs — runtime, skill, plugin, tier, memory record, task, and more.

How to read this page

Terms are grouped by the area they belong to. Each entry is a single sentence — follow the linked concept page when you need the full picture.

Runtime: The GolemCore Bot process — one container, one dashboard, one workspace volume. All other primitives live inside it.
Workspace: The mounted volume at STORAGE_PATH that holds sessions, preferences, skills, memory, traces, and artifacts. Lose the volume, lose all runtime state.
Sandbox: The separate volume at TOOLS_WORKSPACE where shell and filesystem tools operate. Kept distinct from the workspace so tool output cannot corrupt runtime state.
Session: One continuous conversation between a user and the agent, persisted under workspace/sessions/ with its full tool-call history.
Turn: One request/response cycle inside a session: input in, tool loop, response out.
Channel: Any entry point that can start a turn — dashboard chat, Telegram, webhooks, Hive command flows.

Skill: A sticky overlay on an active session: instructions plus optional MCP server and variables. Shapes behavior on an existing capability surface.
Plugin: A capability pack contributed to the runtime at startup — browser, search, mail, voice, RAG. Grows the capability surface itself.
MCP server: An external tool server launched from a skill under the Model Context Protocol. Its tools are exposed alongside native tools for the lifetime of the skill activation.
Tool: A function the agent can call during a tool loop. May come from the core runtime, a plugin, or an MCP server.
Tool loop: The inner cycle inside a turn where the model repeatedly requests tools, the runtime executes them, and the results are fed back until the model stops asking.

Tier: An abstract model role — balanced, smart, coding, deep — resolved to a concrete provider and model through the model router.
Model router: The mapping from tier to concrete model, stored in preferences/model-router.json. Changed without touching the skills or prompts that reference tiers.
Provider: An upstream LLM API (Anthropic, OpenAI, Groq, a local server). Configured once in preferences/llm-providers.json; referenced from the router by name.
Dynamic escalation: Runtime-initiated tier bump when the current model cannot make progress (context too long, repeated tool failure). Always upward, never silent — visible in Sessions.

Memory record: A single structured fact in the memory store. Has a lifecycle: Draft → Confirmed → Updated → Forgotten.
Progressive disclosure: Loading strategy: memory first surfaces short headers, and the agent expands only the records it needs. Keeps context small while preserving depth.
Trace: A detailed, replayable snapshot of a turn — tool calls, inputs, outputs, timings. Inspected from Sessions.
Diary: Auto Mode's append-only log of observations from each goal iteration. Used to carry context across scheduled runs without reloading everything.

Auto Mode: Scheduled, goal-driven execution loop. A goal owns a task list and a diary; a cron trigger advances the next task.
Goal: A long-running objective in Auto Mode with its own task list and diary. Exists independently of any one session.
Task: A unit of work under a goal. State machine: PENDING → IN_PROGRESS → COMPLETED / FAILED / SKIPPED.
Delayed action: A capability loaned to the agent by the runtime to schedule future work — REMIND_LATER, RUN_LATER, NOTIFY_JOB_READY. The runtime owns timing; the agent only requests it.
Webhook hook: A named webhook endpoint with a Mustache template that turns an incoming HTTP payload into an agent prompt.

Hive: Optional fleet orchestrator that coordinates many bots through approvals, lifecycle signals, and shared inspection. Not required for single-runtime deployments.
Dashboard: The web UI served by the runtime on port 8080. Main control surface: Chat, Settings, Scheduler, Sessions, Logs, Skills, Diagnostics, Plugin Marketplace.
Preferences: The stack of JSON files under workspace/preferences/ that hold runtime configuration. The dashboard is the normal editor; the files are the source of truth.

Overview

The shape of the runtime and where each primitive lives.

Troubleshooting

Symptom-to-fix lookup for operational problems.

Overview

What GolemCore is, how the runtime is shaped, and where to go next in the documentation.

User Guide

Quickstart

Install the runtime with one Docker command, open the dashboard, add a provider, and send a first message.