Terms
Every term in this hub.
This hub collects the reliability terms in the Prefactor glossary. The deployment group covers how new agent versions reach production without betting everything on them: canary deployments, blue-green releases, and environment promotion. If your agents currently ship by editing a prompt in production, this group is the argument for stopping.
The versioning group covers knowing what is actually running: version pinning, immutable versions, and the lifecycle states between pending and retired. It sounds bureaucratic until the first incident review where nobody can say which prompt produced the bad output. The failure group covers what happens when things break anyway: retries, fallbacks, timeouts, circuit breakers, and recovery patterns adapted for systems whose failures are often probabilistic rather than crashes.
Drift gets its own group because it is the failure mode unique to this field: an agent that degrades without any deploy, because a model updated upstream or the input distribution moved underneath it. Drift detection is where reliability and evaluation meet, and the entries link across accordingly.
Terms are grouped below in lifecycle order: deployment first, then versioning, then failure and recovery, then drift and degradation. Each entry links to its full definition.
Deployment
Agent Deployment Agent deployment is the process of releasing an agent version into a production or staging environment. Blue-Green Deployment (Agent) Blue-green deployment for AI agents maintains two identical production environments: one live (blue) and one staging the new version (green). Canary Deployment (Agent) A canary deployment for AI agents gradually routes increasing proportions of production traffic to a new agent version while monitoring quality and error metrics. Environment Promotion (Agent) Environment promotion is the controlled process of moving an agent version from development through staging, pre-production, and production environments, with validation gates and approval requirements at each stage.
Versioning
Agent Lifecycle The agent lifecycle is the full sequence of stages an AI agent passes through, from initial design and registration through deployment, active operation, updates, and eventual retirement. Agent Lifecycle Management Agent lifecycle management is the process of governing an agent from registration through deployment, rotation, suspension, and retirement. Agent Versioning Agent versioning is the practice of tagging each release of an agent's model, prompt, tools, and configuration so that any past behavior can be reproduced. Context Versioning Context versioning is the practice of tracking changes to a context schema or context payload format over time. Prompt Version Control Prompt version control is the practice of treating AI prompts as first-class software artefacts, storing them in version control systems, tracking changes with commit messages, enabling branching and merging, and associating each prompt version with its evaluation results.
Failure & recovery
Drift & degradation
More agent reliability & deployment terms
Agent CI/CD Agent CI/CD is the practice of applying continuous integration and continuous delivery principles to AI agent development: automatically testing agents on every change, gating production releases on evaluation results, and automating deployment pipelines for agent code, prompts, and configuration. Agent Configuration Management Agent configuration management is the governed tracking and deployment of all parameters that shape agent behavior, including model selection, temperature, system prompts, tool lists, and policy bindings. Agent Discovery Agent discovery is the process of identifying and cataloguing all AI agents operating within an organisation, including agents that were deployed without central approval. Agent Load Balancing Agent load balancing distributes incoming requests across multiple instances of an agent to optimise throughput and reliability. Agent Onboarding Agent onboarding is the process of registering a new AI agent with the governance system before it is permitted to run in production. Agent-as-a-Service Agent-as-a-Service is a deployment and commercial model where AI agent capabilities are offered over an API, allowing organisations to consume agent workflows without building or operating the underlying infrastructure. AI Agent for Code Generation A code generation agent writes, reviews, or refactors software code based on natural language instructions. AI in Financial Services AI in financial services encompasses the use of AI agents for trading, risk assessment, fraud detection, customer service, and regulatory compliance. AI in Government AI in government uses AI agents for public services, policy analysis, and administrative processes. AI in Legal AI in legal applies AI agents to contract review, legal research, discovery, and compliance analysis. AI Inventory An AI inventory is a comprehensive register of all AI systems, models, and agents deployed or in development within an organisation. AI Liability AI liability refers to the legal responsibility for harm caused by AI systems, including AI agents that take autonomous actions with real-world consequences. AI Operating Model An AI operating model defines how an organisation structures teams, processes, and technology to build, deploy, and govern AI systems at scale. AI Red Teaming AI red teaming is the practice of systematically probing an AI system for vulnerabilities, biases, and failure modes by simulating real-world attacks and edge cases. AI Risk Assessment An AI risk assessment is a systematic evaluation of the potential harms, failure modes, and likelihood of adverse outcomes associated with deploying an AI system. AI ROI AI return on investment measures the business value generated by AI systems relative to the total cost of building, deploying, and governing them. AI Use-Case Registry An AI use-case registry is a centralised inventory of all AI applications and agent workflows deployed or planned within an organisation. Chief AI Officer (CAIO) A Chief AI Officer is the senior executive responsible for AI strategy, governance, and deployment across an organisation. CI/CD for AI CI/CD for AI applies continuous integration and continuous deployment practices to AI agent development: automatically testing, evaluating, and deploying changes to models, prompts, tools, and policies through a governed pipeline. Enterprise AI Strategy An enterprise AI strategy is the organisation-wide plan for how AI will be adopted, governed, and scaled to achieve business objectives. GitOps (Agent) GitOps for AI agents is the practice of managing agent configuration, prompts, policies, and deployment definitions as code in version-controlled repositories, so that the current state of all agents in production can be reproduced from source, every change is reviewed and approved, and rollback is as simple as reverting a commit. Governance by Default Governance by default is the design principle that AI agents are governed from the moment they are created, rather than having governance added retrospectively. Infrastructure as Code (AI) Infrastructure as code for AI defines agent deployment environments, model endpoints, governance integrations, and monitoring pipelines in version-controlled configuration files. Instruction Tuning Instruction tuning is a fine-tuning technique that trains a model to follow natural language instructions more reliably. MITRE ATLAS MITRE ATLAS (Adversarial Threat Landscape for Artificial-Intelligence Systems) is a knowledge base of adversarial tactics, techniques, and case studies for AI systems: analogous to MITRE ATT&CK for traditional cybersecurity. Model Repository A model repository is a versioned store for AI models, their metadata, evaluation results, and deployment history. Prompt Engineering Prompt engineering is the practice of designing and refining the instructions given to a language model to elicit desired behavior. Prompt Fuzzing Prompt fuzzing is an automated testing technique that generates large volumes of varied, unexpected, or adversarial inputs to discover edge cases and vulnerabilities in an AI agent's behavior. Prompt Registry A prompt registry is a versioned store of all system prompts, user prompt templates, and prompt chains used across an organisation's AI agents. Responsible AI Lead A responsible AI lead is the organisational role accountable for defining and operationalising responsible AI principles across AI projects and agent deployments. Responsible AI Principles Responsible AI principles are the ethical commitments an organisation makes about how it will develop and deploy AI systems, covering values such as fairness, transparency, accountability, privacy, and safety. Responsible AI Standard (Microsoft) Microsoft's Responsible AI Standard is an internal framework of principles and requirements for building AI systems responsibly, covering fairness, reliability, privacy, inclusiveness, transparency, and accountability. Shadow Testing (Agent) Shadow testing runs a new agent version in parallel with the production version on real traffic, capturing its outputs without serving them to end users. Shared Services (AI) A shared services model for AI centralises common agent infrastructure, identity management, evaluation harnesses, deployment pipelines, governance controls, and monitoring, and offers them as internal services to product teams. Site Reliability Engineering (AI) Site reliability engineering for AI applies SRE principles, error budgets, service level objectives, toil reduction, and blameless post-mortems, to the operation of AI agent systems.
Questions
Common questions.
What makes agent reliability different from normal software reliability?
Agent failures are often probabilistic rather than crashes: an agent can degrade without any deploy, because a model updated upstream or input data shifted. Reliability practice adds drift detection to the usual deployment discipline.
What is a canary deployment for agents?
Routing a small, growing share of production traffic to a new agent version while watching quality and error metrics, so a bad release burns a fraction of users instead of all of them.