Terminology

ripr uses plain language on its public surfaces (storefront, README, Quickstart, editor) and a precise internal vocabulary in its specs, schemas, JSON, and report artifacts. This page is the bridge between the two, so the internal model stays teachable instead of leaked.

If you only care about the day-to-day language, the plain-language column is all you need. If you are reading JSON, writing fixtures, or working on the analyzer itself, you will want the internal term.

Bridge

Public / first-hour phrase	Internal term	Meaning
changed code where the nearby tests may not actually catch the behavior	seam / finding	the location ripr is worried about
assertion or check that would catch the changed behavior	discriminator	the test evidence ripr is looking for
nearby test to imitate	related test	candidate test context used to suggest the next focused test
the evidence got better after adding a test	movement	before/after static change observed by ripr outcome
top test gap	top actionable seam	the highest-priority seam in the current pilot/report
no focused test gap found	no-actionable-seam	stable status ID emitted when analysis ran cleanly but produced no actionable recommendation
recommended next test	first-useful-action	the advisory target/ripr/reports/first-useful-action.json projection
static classifications: exposed, weak, unrevealed, no static path, unknown	exposed / weakly_exposed / reachable_unrevealed / no_static_path / infection_unknown / propagation_unknown / static_unknown	conservative static-exposure labels — ripr does not use mutation-runtime words such as killed or survived outside calibration reports
ripr's view of how well a behavior is tested	grip (seam-grip class)	seam-native classification across the five RIPR stages, surfaced as SeamGripClass in JSON
identity for a behavior gap (so it does not move when line numbers change)	canonical gap identity	hash of owner / kind / flow sink / missing discriminator / assertion shape used by ledgers, baselines, and gate comparison
PR review summary	pr-review front-panel / report packet index	composed advisory CI artifact summarizing PR guidance, first useful action, assistant proof, ledger, baseline, gate, calibration, coverage/grip, and receipt
agent proof status	assistant-loop health	advisory summary of existing assistant proof reports
receipt for a focused test	agent receipt	provenance plus bounded next-action guidance for one focused test
under-the-hood model	RIPR: Reachability, Infection, Propagation, Revealability	the four stages ripr evaluates statically per probe / seam

Where the precise vocabulary lives

The internal terms are stable in:

• Output schema — JSON contracts for findings, repo exposure, agent packets, and calibration reports.
• Static exposure model — the domain model.
• Specs — RIPR-SPEC documents pin behavior; RIPR-SPEC-0021 pins the seam-native evidence_record projection.
• Capability matrix and Metrics.
• Implementation campaigns — full build history of the surfaces in the current product.

Renaming the public-facing copy of any of these terms does not rename the JSON field, status ID, command, schema, or spec it maps to.

When to use which

Use the plain-language column in:

• the marketplace storefront and editors/vscode/README.md,
• the root README.md first screen,
• Quickstart,
• Editor extension status copy and first-use steps,
• generated CI advisory summary headings (work in progress).

Use the internal column in:

• docs/specs/**,
• docs/OUTPUT_SCHEMA.md,
• metrics, fixtures, and report schemas,
• implementation docs and campaign notes,
• CHANGELOG.md entries that describe internal contracts.

Linking this page from a spec is unnecessary — specs are already the authority on internal terms.