Glossary

Shared terms currently used across /unitt.wiki.

Platform Terms

Term	Meaning
/unitt.wiki	The documentation site for Unitt platform concepts, developer pages, Unitt Market pages, and reference material.
Agent	A runtime system that uses objectives, tools, connectors, skills, memory, policies, and governance controls to execute work.
Agent Fabric	A collaborative runtime layer where multiple specialized agents coordinate through shared objectives, workflows, memory, and execution patterns.
Agent Runtime	The execution environment where a Unitt evaluates objectives, applies policies, uses tools, calls connectors, and coordinates workflows.
Assembly	The primary developer section for defining the core building blocks of an agent, including Core, Objectives, Patterns, Connectors, Tools, and Skills.
Coming Soon	A banner used on placeholder pages that are planned but not fully documented yet.
Developers	The documentation area for building, configuring, validating, and operating Unitt runtime systems.
Emergence	The platform area for advanced runtime composition using memory architectures, subunits, adaptive state patterns, simulation, and optimization.
Fabric	The platform area for coordinated multi-agent runtime systems, shared workflows, distributed execution, validation, and publishing.
Market	The documentation area for publishing, licensing, deploying, and operating reusable agentic systems.
Unitt Market	The distribution model for publishing fully tested agentic systems as deployable runtime fabrics.
Runtime	The active execution lifecycle where an agent reasons, validates, calls tools, uses connectors, and advances workflows.
Runtime Fabric	A reusable operational system composed of agents, workflows, tools, connectors, memory, policies, and validation logic.
Runtime Session	A single operational execution context where runtime state, objectives, memory, tools, and governance rules are evaluated.
Runtime Workflow	A structured sequence, graph, or process used by an agent to complete objectives during execution.
Unit	A runtime entity or agent unit; used in some Core content as a general form of Unitt.
Unitt	The platform-specific agent or runtime unit configured through Core, Objectives, Patterns, Connectors, Tools, and Skills.

Core And Governance

Term	Meaning
Access Scope	The allowed range of systems, data, credentials, or actions that a connector, tool, vault, or runtime may use.
Agent Initialization	The setup phase where Core files establish identity, rules, policies, and governance before execution begins.
Audit Trail	Recorded execution evidence used for observability, accountability, compliance, and governance review.
Authority Scope	The boundary that defines what an agent is allowed to decide, access, or execute.
Behavioral Boundaries	Rules, policies, and constraints that limit unsafe, excessive, unauthorized, or misaligned runtime behavior.
Budget Limit	A financial, token, API, or runtime cost boundary enforced during execution.
Confidence Threshold	A minimum confidence level required before a runtime action may proceed without review or escalation.
Core	The Assembly section that defines Identity, Rules, Policies, and Governance for a Unitt before runtime execution begins.
Core File	A markdown file such as identity.md, rules.md, policy.md, or governance.md that configures part of the Core workload.
Governance	The control framework that enforces policies, validates behavior, manages escalation, records audit trails, and preserves oversight.
governance.md	The Core file that defines supervision, enforcement, auditability, escalation, and operational oversight models.
Human Review	A manual approval or escalation step used when runtime actions require oversight before continuing.
Identity	The foundational operational character, behavioral posture, specialization, authority scope, and interaction model of a Unitt.
identity.md	The Core file that defines the persistent operational profile and behavioral framing of a Unitt.
Operational Profile	The persistent identity and behavior model that stabilizes how a Unitt reasons, communicates, and executes.
Persona Configuration	Minimal or specialized identity configuration used to shape how a simpler Unitt behaves.
Policies	Conditional governance and authorization logic used to dynamically regulate runtime behavior.
policy.md	The Core file that defines authorization conditions, runtime gates, safety controls, and adaptive policy behavior.
Policy Engine	A governance component that evaluates policies against runtime context, actions, tools, connectors, and risks.
Recall Period	A strict interval or condition defining when an external database, API, CLI tool, or connector should be recalled.
Rules	Static or persistent behavioral constraints used to guide execution patterns, tool validation, and runtime behavior.
rules.md	The Core file that defines rule-based boundaries and execution constraints.
Tool Authorization	The governance decision that determines whether a runtime may use a specific tool.

Objectives And Patterns

Term	Meaning
Constraint	A maximum operational boundary that limits what a Unitt may attempt while pursuing objectives.
Constraint Prompt	A prompt that asks the system to define operational limits such as approval requirements, budgets, allowed systems, and stop conditions.
Draft Email	An example workflow output generated before human approval or final delivery.
Execution Pattern	A reusable runtime model that defines ordered, parallel, conditional, or iterative workflow behavior.
Example Outcome Constraints	The listed output constraints generated from a constraint prompt.
Example Sub-Objectives	A staged breakdown of smaller goals used to complete a primary objective.
Example Workflow Description	Plain-language text used to generate a workflow graph.
Human Approval	A workflow condition requiring a person to approve an action before the runtime proceeds.
Mermaid	The diagram syntax used to render flowcharts and workflow graphs in the wiki.
Objective	The intended outcome that a Unitt is expected to achieve and validate during runtime execution.
Pattern	The internal workflow structure that determines how a Unitt executes objectives.
Pattern Builder	The interface or prompt process that converts plain-language workflow descriptions into structured execution graphs.
Primary Objective	The main mission or expected process output for a Unitt.
Success Condition	A measurable result or validation requirement used to determine whether a step, sub-objective, or objective succeeded.
Sub-Objective	A smaller staged execution goal that contributes to completion of the primary objective.
Validation Checkpoint	A workflow point where output, progress, confidence, or operational correctness is checked.
Workflow Graph	A structured graph showing workflow steps, dependencies, conditions, and completion paths.
Workflow Stage	A single step or state within a larger workflow graph or execution process.

Skills And Tools

Term	Meaning
Additional Common Runtime Tools	Optional runtime tools that may be added as workflows expand or operational requirements evolve.
Anatomy of a Tool	The concept that a tool is defined by a JSON declaration and a markdown usage file.
API Wrapper	A tool or connector layer that exposes API behavior through a controlled runtime interface.
Audit	A default core tool that stores, validates, and reviews internal runtime patterns, execution traces, and governance signals.
Auth	A default core tool that retrieves scoped sessions, credentials, and vault-authorized access tokens.
Browser	A runtime tool for controlled web interaction, browsing, scraping, validation, research, or automation.
CI/CD Orchestration	A complex skill use case for coordinating continuous integration and continuous deployment workflows.
CLI	A command-line interface used by shell tools, local binaries, infrastructure tooling, and runtime utilities.
Comms	A default core tool that bridges runtime communication to external systems, operators, workflows, and notification systems.
Complex Skill	A multi-stage capability that combines workflows, tools, connectors, objectives, validation logic, policies, and runtime behaviors.
Complex Skill Structure	The internal composition of a complex skill, including markdown skill files, workflow graphs, tool dependencies, policies, and constraints.
Default Core Tools	The baseline tools available to support authentication, communication, model access, file operations, and auditing.
File	A default core tool that provides controlled local file system operations such as search, read, write, edit, and validation.
GitHub-Based Utility	A tool imported from GitHub that can be built and attached to the runtime.
JSON Declaration	The structured runtime definition that registers a tool and defines commands, inputs, outputs, permissions, and execution rules.
Local Binary	A local executable file registered as a runtime tool.
Memory	A runtime tool and Emergence concept for context retrieval, embeddings, summaries, recall, and state persistence.
MLOps Evaluation Pipeline	A complex skill use case for evaluating machine learning or model operations workflows.
Model	A default core tool that acts as the gateway to LLM providers and model runtimes.
Monitor	A runtime tool that collects telemetry, logs, metrics, health states, budget usage, and diagnostics.
Runtime Activation	The process of loading skills, tools, connectors, policies, memory, or sub-skills when a workflow requires them.
Runtime Philosophy	The principle that agents should begin with the minimum required tools and add capabilities only as needed.
Runtime Skill Usage	The runtime process for deciding which skills should be loaded or activated during execution.
Sandbox	A runtime tool for safely inspecting, testing, running, or operating isolated workloads outside production systems.
Schedular	A runtime tool that schedules, queues, retries, and monitors jobs, workflow stages, delayed actions, and timed tasks.
Script	A buildable or executable runtime dependency used as a tool.
Search	A runtime tool for indexed local or remote search across files, memory, APIs, documentation, or connected systems.
Shell	A runtime tool that executes approved CLI commands and local automation tasks within controlled boundaries.
Skill	A reusable markdown knowledge module that guides when, how, and why operational behavior should be used.
Skill Registration	The process of registering focused markdown files that describe operational capabilities or behavioral specializations.
Tool	An executable capability that allows the runtime to interact with systems, process data, automate work, or perform specialized tasks.
Tool Declaration	The structured JSON definition used to register, validate, expose, and control tool access.
Tool Management	The process for uploading binaries, registering executables, importing utilities, and attaching tools to the runtime.
Tool Usage Markdown	The markdown file that explains when, how, where, and why an agent should use a tool.
Tool Validation	The process of testing how the model interprets a tool call, generated arguments, permissions, outputs, errors, and missing information.
Uploaded Executable	A binary or executable provided by a developer and attached to the runtime as a tool.
Validation	A runtime tool for schema checks, output validation, policy verification, confidence scoring, and execution integrity checks.

Connectors

Term	Meaning
Access Token	A scoped credential used by the runtime to access an external system or connector.
Authentication Method	The mechanism used by a connector to prove identity and access an external system.
Browser & Automation Connectors	Connectors used for browser control, scraping, testing, validation, and runtime interaction.
CLI-Based Runtime Connectors	Connectors that interact directly with local binaries, infrastructure tooling, automation frameworks, and runtime utilities.
Cloud & Infrastructure Connectors	Connectors for infrastructure orchestration, deployment, storage, networking, and cloud automation.
Common Connectors	The standard categories of external systems an agent may interact with during runtime execution.
Communication Connectors	Connectors used for messaging, alerts, workflow coordination, approvals, and operator communication.
Connector	An external system, service, API, database, application, endpoint, or platform that a Unitt may access during execution.
Connector Definition	The structured description of a connector's target system, endpoint, authentication, permission scope, usage, and limits.
Connector Testing	The validation step that verifies credentials, reachability, permissions, and operational status before runtime use.
Credential Vault	A stored credential collection that can be associated with an agent or selected per connector.
Credentials	Secrets, tokens, sessions, or access details used to authenticate connector and tool access.
Data & Storage Connectors	Connectors used for structured data, caches, queues, object storage, search indexes, graphs, and vector databases.
Endpoint	A specific service address, API route, database connection, CLI target, or external system entry point.
External Dependency	Any connector, service, endpoint, credential, tool, or workflow stage outside the core runtime that execution depends on.
External Platform Connectors	Connectors to third-party platforms such as social, source control, documentation, and task management services.
Operational Limit	A connector-specific boundary defining usage frequency, scope, permissions, retry behavior, or failure handling.
Permission Scope	The permitted actions and data access granted to a connector, credential, tool, or runtime.
Search Vaults	The action used to associate an existing credential vault or individual credential with an agent.
Secrets	Sensitive credential material such as tokens, passwords, keys, or scoped authentication data.
System & Runtime Connectors	Connectors used for remote execution, APIs, webhooks, local binaries, and runtime utilities.
User Vault	The initial vault where external connection references, credentials, secrets, scopes, and boundaries are organized.
Vault	A credential storage boundary used to manage access tokens, secrets, credentials, and permission scopes.
Webhook	An event endpoint used to trigger workflows, ingest external events, or coordinate distributed runtime actions.

Connector Examples

Term	Meaning
AWS	A cloud connector for infrastructure orchestration, S3 storage, ECS/EKS deployments, IAM access, and automation.
Azure	A cloud connector for enterprise infrastructure, authentication systems, deployment orchestration, and automation.
Cloudflare	A connector for DNS management, Workers deployment, edge networking, and security workflows.
Discord	A communication connector for community communication, notifications, automation, and runtime interaction.
Docker	A connector for container management, image builds, runtime execution, and deployment automation.
Elasticsearch / OpenSearch	Search and analytics connectors used for indexing, observability, operational analytics, and runtime queries.
Email / SMTP	A communication connector for outbound messages, notifications, reports, approvals, and automated communication workflows.
GitHub	A source control connector for repository access, pull requests, CI/CD workflows, code analysis, and automation pipelines.
GitLab	A source control connector for repository management, deployment automation, and workflow orchestration.
Google Cloud	A cloud connector for infrastructure, storage, deployment workflows, and runtime services.
Jira	A task management connector for tickets, sprint coordination, workflow tracking, and operational management.
Kubernetes	An infrastructure connector for cluster orchestration, workload deployment, scaling, and runtime management.
LinkedIn	A social and business connector for prospecting, professional outreach, company analysis, and business intelligence workflows.
Microsoft Teams	A communication connector for enterprise messaging, approval routing, coordination, and reporting.
Neo4j / Graph Databases	Graph connectors for relationship mapping, memory graphs, workflow dependency analysis, and reasoning systems.
Notion	A documentation connector for workflow state, operational knowledge, and collaborative runtime memory.
Playwright	A browser automation connector for validation workflows, scraping, testing, and runtime interaction.
PostgreSQL	A database connector for structured runtime data, workflow persistence, operational storage, and audit systems.
Puppeteer	A browser automation connector for web automation, scraping, validation, and runtime browsing tasks.
Reddit	A social connector for community analysis, sentiment tracking, research, and monitoring workflows.
Redis	A data connector for caching, queues, distributed state coordination, and transient runtime storage.
REST APIs	System connectors for external service integration, runtime communication, and operational automation.
S3-Compatible Storage	Object storage for artifacts, runtime snapshots, logs, and workflow assets.
Selenium	A browser automation connector for browser control, testing, and workflow interaction systems.
Slack	A communication connector for runtime messaging, alerts, workflow coordination, approvals, and operator communication.
SSH	A system connector for remote execution, infrastructure management, deployment workflows, and maintenance.
Telegram	A communication connector for lightweight messaging, alerts, notifications, and command interaction.
Twilio	A communication connector for SMS, voice communication, verification flows, and notifications.
Vector Databases	Storage connectors for embeddings, semantic search, memory recall, and retrieval-augmented workflows.
X / Twitter	A social connector for posting, monitoring, trend analysis, and engagement workflows.

Emergence And Fabric

Term	Meaning
Adaptive Runtime Strategy	A runtime strategy that changes based on objectives, state, workflow outcomes, or operational context.
Adaptive State-Management Pattern	A strategy for selecting and refining state behavior for specific operational outcomes.
Agentic Execution Strategy	A reasoning and execution approach used by an agentic runtime to complete complex work.
Consensus Validation	A collaborative Fabric pattern where multiple agents or reasoning paths validate an outcome.
Data	The Fabric page for data structures, inputs, outputs, persistence, and movement patterns.
Delegated Sub-Workflow	A workflow segment assigned to another agent or subunit inside a broader execution system.
Distributed Execution Pattern	A Fabric execution model where multiple agents or runtime components coordinate work across a system.
Flow	The Fabric page for execution flow, orchestration paths, dependency handling, and runtime coordination.
Genetic Optimization System	An optimization method referenced in Emergence for mutating and evolving runtime configurations over time.
Memory Architecture	A design for context retrieval, summaries, embeddings, recall, and persistence used by an agent.
Planner-Executor System	A Fabric coordination pattern where planning and execution roles are separated across runtime components.
Publish	The Fabric page for release checks, publishing workflows, distribution paths, and operational handoff.
Reflection Loop	A runtime pattern where an agent reviews, critiques, or improves its own intermediate outputs.
Setup	The Fabric page for configuration, environment setup, and initialization steps.
Simulation Environment	A validation environment used to test runtime systems, outcomes, mutations, and operational behavior.
State	The Emergence page for runtime state, lifecycle transitions, continuity, and persistence.
Subunit	A component runtime inside Emergence where complex runtimes, workflows, skills, and reasoning systems are stacked together.
Test	The Fabric page for validation strategy, fixtures, quality checks, and runtime verification.
WorldSim	The Emergence page for simulation behavior, environment modeling, scenario testing, and operational forecasting.

Market And Consulting

Term	Meaning
Consulting	Unitt Market support for designing, building, validating, deploying, and scaling enterprise-grade agentic systems.
Deployment & Licensing	The Unitt Market model for deploying systems under hosted, managed, shared, private, or transferable licensing terms.
Enterprise-Grade Agentic System	A production-oriented agentic system designed for reliability, security, performance, governance, and measurable outcomes.
Estimated Token Cost	A projected runtime cost based on expected token usage for a published system.
Execution Consistency	An agent market metric describing how reliably a system repeats validated operational behavior.
Fully Transferable Operational Fabric	An agent market system that can be transferred and deployed into the purchaser's infrastructure and vault systems.
Hosted Runtime	An agent market deployment model where the runtime is operated as a hosted system.
Infrastructure Requirement	A required cloud, container, connector, vault, tool, or runtime dependency needed before deployment.
Listing	The Unitt Market page for listing requirements, metadata, and publishing guidance.
Managed Deployment	An agent market deployment model where operation, setup, or infrastructure may be managed on behalf of the user.
Objective Success Rate	An agent market metric for how often a system achieves its intended objectives.
Operational Metric	A measurable agent market signal such as success rate, consistency, runtime pattern quality, simulation outcome, infrastructure need, or cost.
Operational Planning	Consulting work that maps business objectives into runtime architecture, workflows, governance, and deployment paths.
Ownership	The Unitt Market page for ownership models, responsibilities, and governance notes.
Private Ownership License	A licensing model where the purchaser receives ownership rights to operate a runtime system.
Private Ownership Model	An agent market model where the original developer earns 80% of the initial purchase price and 5% of ongoing token consumption costs.
Production-Ready Agentic System	A validated, governable agentic system prepared for operational deployment.
Project Discovery	Consulting discovery work used to understand goals, requirements, constraints, systems, risks, and expected outcomes.
Shared Usage License	An agent market license allowing use of a runtime system under shared or limited usage terms.
Token Consumption Cost	Ongoing cost generated by token usage in a deployed Unitt Market system.
Usage	The Unitt Market page for usage flows, user expectations, and operational notes.