# NovusLens: the core concepts Audience: Business sponsors, executives, team owners and readers new to the Concepts documentation ## s00 · 1 · What the evidence supports · Read the map, measures and limits What does a NovusLens figure tell you about your organisation? The Concepts pages explain how to read its structure, the evidence behind it and who answers for decisions. This overview follows those ideas in five short chapters. Start with evidence, then the organisation model, the measures, accountability and the related pages. Sources: - apps/docs/content/concepts/index.mdx - apps/docs/content/concepts/epistemics.mdx ## s01 · 1 · What the evidence supports · Measure the present NovusLens measures current structure as the evidence changes. It does not forecast business results or rate individual performance. A what-if recalculates a measure for a proposed structure. Its result is indicative: it describes that structure under the stated assumptions. It does not promise a delivery date, a cost saving or a customer outcome. Sources: - apps/docs/content/concepts/epistemics.mdx - apps/docs/content/concepts/car.mdx ## s02 · 1 · What the evidence supports · Read the source and its date Each claim carries one of three evidence labels. Told means someone stated it. Shown means a document supports it. Measured means a connected source recorded an observation within a stated period. These labels describe support, not truth. Check the source, its coverage and any disagreement. When an observation expires, it leaves the current picture and stays in history. Sources: - apps/docs/content/concepts/evidence-ladder.mdx ## s03 · 2 · The organisation model · Units, owners and dependencies A twin is NovusLens’s model of how your organisation is wired. It is only as current as its inputs. Each unit, called a cell, represents a team, service, supplier or AI agent. The ownership tree shows containment. Dependency links show where work relies on another unit. Business domains describe the work and can span several teams. Sources: - apps/docs/content/concepts/organisation-model.mdx ## s04 · 2 · The organisation model · Written accounts and findings The meaning layer holds the organisation’s procedures, definitions and other written accounts beside its structure. Answers can cite those accounts. Findings point to gaps or differences in the records, such as a missing accountable owner or a disputed handoff. Open the supporting sources before interpreting a finding. An empty findings list does not establish that nothing is wrong. Sources: - apps/docs/content/concepts/meaning.mdx - apps/docs/content/concepts/findings.mdx ## s05 · 3 · The measures · Three separate structural readings Complexity, Agility and Recoverability are abbreviated as C/A/R. Complexity describes the structure a unit manages, including its systems and the changes coordinated across them. Agility describes how readily it changes safely, through change lead times and failures. Recoverability describes restoration after failure, through restoration times and tested recovery. NovusLens keeps the three separate. It does not blend them into one score. Sources: - apps/docs/content/concepts/car.mdx ## s06 · 3 · The measures · A score needs three things A score needs a calibration, a scope and current evidence. The calibration is the organisation’s documented rule for converting observations into an indicator. Scope selects the unit’s own operations or includes its owned descendants. External suppliers stay outside that ownership sum. Current evidence covers the relevant observation period. Recording a calibration makes its assumptions visible. It does not make them correct. Sources: - apps/docs/content/concepts/car.mdx ## s07 · 3 · The measures · Unknown is not zero An available measure has a current calibrated claim supporting it. Unknown means that current calibrated evidence is missing for the scope. Conflict means current claims have not been reconciled. Read the reason before using the figure. Compare scores only where calibration, scope and observation periods match. A parent’s score is recalculated from its records, never averaged from its children’s scores. Sources: - apps/docs/content/concepts/car.mdx ## s08 · 3 · The measures · Find the heaviest dependency chain Critical path length, or CPL, identifies the dependency chain with the largest total coordination weight. It uses index points, not hours or money. Read the coverage beside it: some links can still rely on stated accounts. Unknown links are excluded and counted, so incomplete coverage can hide a heavier chain. The result points to dependencies worth examining with their owners. Sources: - apps/docs/content/concepts/cpl.mdx ## s09 · 4 · Accountability and the boundary · Three questions for each AI agent Capability, Authority and Responsibility answer three separate questions about an AI agent. Capability records what it declares it can do. Authority records what it is allowed to do. Responsibility names who answers for the result. NovusLens compares these declarations with activity reported by the organisation’s systems. Gaps become findings for review. These accountability questions are separate from the three structural measures. Sources: - apps/docs/content/concepts/agentic-accountability.mdx ## s10 · 4 · Accountability and the boundary · People own decisions and controls The operational owner runs a unit day to day. The accountable owner answers for its outcome. NovusLens keeps those records separate. It maps, surfaces findings and advises. The customer’s own systems control its agents. The Layer-A/B boundary page explains this division. NovusLens does not run or stop customer agents, and it never issues a compliance verdict. Sources: - apps/docs/content/concepts/agentic-accountability.mdx - apps/docs/content/concepts/layer-boundary.mdx ## s11 · 5 · Where to read next · Several regional models, one view Federation lets regional twins appear together in a group view. Each member contributes summary figures, alongside recorded cross-region dependencies. The group view does not read a member’s raw records. A group-wide critical path stays partial while any cross-region coupling is unknown. Read Federation and groups for the sharing boundaries and the current limits of regional signature checks. Sources: - apps/docs/content/concepts/federation.mdx ## s12 · 5 · Where to read next · Activity records and privacy The activity log records the product’s own use, including recorded reads and changes. It does not show work outside NovusLens or rate anyone’s performance. The privacy floor keeps small personnel counts from identifying people. A key-person finding reports fewer than five mapped people without names or an exact count. Read these two pages for the access rules and reporting limits. Sources: - apps/docs/content/concepts/activity-log.mdx - apps/docs/content/concepts/privacy-floor.mdx ## s13 · 5 · Where to read next · Choose the page for your question Use the Concepts overview to choose your next page. For a figure, read the three measures or critical path length, then inspect its evidence. For an ownership question, read the organisation model and agentic accountability. For a finding, open its sources and read its current support. Keep the scope and the evidence beside any conclusion you take into a decision. Sources: - apps/docs/content/concepts/index.mdx - apps/docs/content/concepts/car.mdx - apps/docs/content/concepts/cpl.mdx - apps/docs/content/concepts/organisation-model.mdx - apps/docs/content/concepts/agentic-accountability.mdx - apps/docs/content/concepts/findings.mdx - apps/docs/content/concepts/epistemics.mdx