TEA glossary

Argument and evidenceCopy Anchor Link

• Argument: The stated reasoning that connects a goal, its supporting claims and the evidence offered for them.
• Assurance: Work that builds justified confidence in a system through questions, actions, argument and evidence.
• Assurance case: A structured, reviewable argument about a system in a stated context.
• Claim: A proposition about a system or project that can be supported or challenged.
• Goal: The broad proposition at the top of a case. The editor creates G1 when a new case is made.
• Strategy: An explanation of how a goal or claim is broken into parts that can be supported.
• Property claim: A narrower proposition about the system or project that a reviewer can test or challenge.
• Evidence: An artefact or result offered for a particular claim. Its relevance, reliability and sufficiency still need assessment.
• Evidence quality: A judgement about whether evidence is relevant to a claim, comes from a credible source and is sufficient for the proposed conclusion.
• Relevance: The degree to which evidence addresses the particular claim to which it is linked.
• Reliability: The trustworthiness of an evidence source, method and record.
• Sufficiency: Whether the available evidence is enough for the claim and its stated scope.
• Context: A definition or boundary needed to understand where a claim applies.
• Assumption: A condition taken as given in an argument and open to challenge if it fails.
• Justification: A reason for including a claim, choice or decomposition in the argument.
• Defeater: An element used to challenge another element. Its name uses C before its normal type prefix, such as CP1.
• Away goal: An element that cites a goal or other valid target in a referenced case.
• Module: A reference to another case used as part of an argument.
• Assertion status: The author's declared position on an assertion: Asserted, Needs support, Assumed, Axiomatic or Defeated. As cited is reserved and is not computed automatically in this build.

Project reasoningCopy Anchor Link

• Critical reflection: Examination of a project's assumptions, possible harms, argument and evidence, repeated as the project develops.
• Deliberation: Careful discussion of options, values and trade-offs with people who bring relevant perspectives.
• Stakeholder engagement: Involving people who use, build, govern or may be affected by a system while its questions and choices can still change.
• Ethical principle: A broad aim, such as fairness, that starts deliberation but needs project-specific interpretation.
• Fairness: A concern with how a system affects people's rights, interests and outcomes, specified for a particular use setting.
• Explainability: The ability to give a useful account of system behaviour to the people who need it, considering the wider use setting as well as the model.
• Accountability: Clear responsibility for project decisions and a record that lets others understand and challenge them.
• Proportionality: Giving inquiry and evidence work attention in relation to a project's likely effects and use setting.
• Specification: Turning a broad principle into a statement precise enough to guide decisions for a particular system.
• Operationalisation: Identifying, weighing, specifying, revising and implementing principles in project activity.
• Justified trust: Confidence grounded in inspectable reasons and evidence, with limits stated.

Platform collaborationCopy Anchor Link

• Stakeholder: A person or group who uses, builds, governs or may be affected by the system.
• Trustworthy and Ethical Assurance (TEA): The practice of building a reviewable argument, with evidence and stakeholder reasoning, for a system in context.
• Working case: The editable case available to people and teams with case permission.
• Team role: A person's membership role within a team; it does not automatically confer case access.
• Case permission: A direct or team grant to view, comment, edit or administer a working case.
• Comment: Discussion attached to a case or element, kept outside the public published snapshot.
• Live update: A case event that prompts another open editor view to refresh after a change.
• Public snapshot: A published version available on Discover under a slug, separate from the working case.
• Discover: The public area where published case versions can be found and read.
• Slug: The readable path segment used for a published case page.
• Backup: A separate exported copy of case data, such as JSON saved to Google Drive.
• Machine integration: A registered external service that uses scoped API tokens and case grants.
• Health evidence: An append-only, hash-chained record of one runtime check result against a property claim, with a metric, value, threshold, verdict and provenance.
• API token: A secret beginning teap_ that a machine client sends as a bearer credential and that TEA shows once on issue.
• Scope: A named kind of machine action allowed to an integration; case permission still controls which case it can use.
• Trash: The recovery area for deleted cases before the 30-day purge cutoff.

For the controls behind these terms, see platform basics and element types.