SENAR Reference: Complete Glossary
Alphabetical listing of all SENAR terms. Core terms are defined normatively in SENAR Standard Section 3.
| Term | Definition |
|---|---|
| AI Agent | Software system powered by LLM that generates engineering artifacts under human direction |
| AI Model Provider | External service providing AI Agent capabilities (e.g., Anthropic, OpenAI, Google); de facto supplier (Standard 3.25) |
| AI Model Version | Specific version/generation of an AI model; metric baselines are version-dependent (Standard 3.26) |
| Checkpoint | Context preservation action during a Session |
| Context | Information provided to AI Agent: goal, AC, constraints, knowledge, traceability |
| Context Architect | Responsibility: designs requirements as structured AI input, manages traceability |
| Cost per Task | Metric: total cost / tasks done, segmented by complexity |
| Cost Predictability | Metric: actual cost / planned cost for an Increment |
| Cycle Time | Time from Task start to completion (started_at → completed_at) |
| Dead End | Documented failed approach with reason for abandonment |
| Defect Escape Rate | Metric: post_done_defects / done × 100% — defect tasks linked to a completed parent, over Tasks completed (Standard 9.1) |
| Delivery Review | Ceremony: demonstrate software to stakeholders |
| Exploration | Time-bounded investigation without full Task formality |
| Federation | Coordination mechanism for multiple Supervisor+AI Pairs |
| Federation Sync | Ceremony: coordinate multiple Pairs on dependencies |
| First-Pass Success Rate | Metric: % Tasks completed correctly in one cycle |
| Flow Manager | Responsibility: session rhythm, cost tracking, flow metrics |
| Gate Bypass | Documented exception to proceed past a Quality Gate. Admission of a gate’s declared effect without a positive verdict is a bypass whether or not it was intended as one, and is recorded as one (Standard 8.6(h), 8.6(j)) |
| Gate Properties | The five properties a check has to have to be a Quality Gate: declared prevented effect, preventive placement, correspondence of the verdict to state, binding to a measured digest, fail-closed. Obligation levels scale by configuration (Standard 8.6) |
| Increment | Scope-bounded batch of work with objectives and budget |
| Increment Planning | Ceremony: define objectives, tasks, budget, risks |
| Increment Retrospective | Ceremony: quantitative review of Increment metrics |
| Knowledge Capture Rate | Metric: knowledge entries / tasks done |
| Knowledge Engineer | Responsibility: capture, curate, maintain knowledge base |
| Knowledge Entry | Documented decision, pattern, gotcha, or dead end |
| Lead Time | Time from Task creation to completion |
| Manual Intervention Rate | Metric: % Tasks with a Gate Bypass recorded under Standard 8.6(j). Derived from records, not from self-report (Standard 9.2) |
| Maturity Level | Organization’s SENAR adoption depth (L1–L5) |
| Quality at Input | Principle: defects in requirements cascade to all downstream artifacts; quality is built at input, not checked at output |
| Quality Gate | Enforcement point that prevents a declared effect until its verdict is positive. Meeting the criteria of 8.1–8.5 is not sufficient: a check lacking a property required at the applicable configuration is not a Quality Gate (Standard 8.6) |
| Prevented effect | The specific change to the system, or to the record of work, that does not occur while a gate’s verdict is negative. Declared per gate in Standard 8.1–8.5; required by 8.6(a) |
| Quality Sweep | Ceremony: periodic comprehensive codebase/KB audit |
| Requirement | Documented, verifiable statement of need at BR, SR, or TR level (Standard 3.17) |
| Requirement — Business (BR) | Stakeholder-level need in business terms; source of all downstream requirements (Standard 3.18) |
| Requirement — System (SR) | System-level capability derived from BR; what the system must do (Standard 3.19) |
| Requirement — Task (TR) | Implementation-level requirement = Task goal + acceptance criteria (Standard 3.20) |
| Requirement Hierarchy | Decomposition chain: BR → SR → TR; depth scales by complexity (Standard 3.21) |
| Requirement Library | Managed repository of verified, reusable requirements stored in Knowledge Base (Team+) |
| Requirement Pattern | Reusable AC template for common task types (CRUD, migration, UI, integration) |
| Session | Time-bounded period of supervised AI work |
| Session End | Ceremony: capture metrics, write handoff, record knowledge |
| Session Start | Ceremony: load context, select tasks, verify environment |
| Story | Intermediate grouping of Tasks as stakeholder-visible deliverable |
| Supervisor | Human who directs AI agents, verifies output, enforces gates. Production of artifacts is AI-directed; direct modification by the Supervisor is a recorded Gate Bypass (Standard 3.2, 4.1) |
| Supervisor+AI Pair | Fundamental production unit: one Supervisor + AI Agent(s) |
| Task | Atomic unit of tracked work with goal and acceptance criteria |
| Throughput | Metric: tasks completed per Session |
| Traceability | Bidirectional linking: every TR → BR upward; every BR → TR(s) downward (Standard 3.24) |
| Value Stream | End-to-end flow from client request to delivered software |
| Verification Engineer | Responsibility: audits AI output for correctness and security |
| Work Type | Functional category of a Task (dev, arch, QA, docs) |
| WSJF | Prioritization: Cost of Delay / Job Size (from SAFe) |