diff --git a/packages/cli/AGENTS.md b/packages/cli/AGENTS.md
index 8d8ed44e1..4a23eba7e 100644
--- a/packages/cli/AGENTS.md
+++ b/packages/cli/AGENTS.md
@@ -420,7 +420,7 @@ takes:
 | Comma-separated fixed values (`--scope`, `--features`) | `choiceListParser(values)` (or `(values, parse)`) | the next item of the list, leaving out the items already in it (`org:read,org:w` → `org:read,org:write`) |
 | Platform | `parsePlatform` | platform identifiers, common ones first |
 | Something that already exists in Sentry, as a flag value (`--environment`, `--release`, `--team`, `--owner`, `--integration`, `--app`, `--dsn`, `--workflow`, `--project`, `issue tags <key>`) | `parseEnvironment`, `parseEnvironmentList`, `parseRelease`, `parseTeam`, `parseOwner` (`team:<id>`), `parseIntegration`, `parseSentryApp`, `parseAppLinkField` (`name=value`), `parseDsn`, `parseWorkflowList`, `parseProjectSlug`, `parseProjectList`, `parseTagKey` | cached entities of the org the command's target names, else of every org, each once; the `*List` parsers complete the last item of a comma-separated list |
-| Field or aggregate of a dataset (`explore --field`/`--sort`, `dashboard widget --group-by`/`--query`/`--sort`, `alert metrics --aggregate`) | `fieldParser`, `fieldOrAggregateParser`, `aggregateParser` (`fn(field)`), or `aggregateShorthandParser` (`fn:field`), given a `DatasetCompletion`: what completes on each dataset (start from `DATASET_FIELDS`), the command's `--dataset` aliases, and its default dataset | the fields of the dataset the `--dataset` typed names (else the default; every dataset when there is none): cached attribute keys of spans, logs, and trace metrics, cached issue tag keys for errors, or a fixed set (replays); then its aggregate functions, from `aggregateFunctionsOf` in `types/dashboard.ts`, as `fn()`, and `fn(field)` once a name is typed in full (`p95` → `p95(span.duration)`). A word starting with `-` completes flag names, so a descending sort's `-` is typed after completing |
+| Field or aggregate of a dataset (`explore --field`/`--sort`, `dashboard widget --group-by`/`--query`/`--sort`, `alert metrics --aggregate`) | `fieldParser`, `fieldOrAggregateParser`, `aggregateParser` (`fn(field)`), or `aggregateShorthandParser` (`fn:field`), given a `DatasetCompletion`: what completes on each dataset (start from `DATASET_FIELDS`), the command's `--dataset` aliases, and its default dataset | the fields of the dataset the `--dataset` typed names (else the default; every dataset when there is none): cached attribute keys of spans, logs, and trace metrics, cached issue tag keys for errors, or a fixed set (replays), matched one `.` segment at a time (`span.d` finds `span.duration`, never `span.op`); then its whole aggregates (`aggregates`), and its aggregate functions as `fn()`, and `fn(field)` once a name is typed in full (`p95` → `p95(span.duration)`). Widgets take the functions of `aggregateFunctionsOf` in `types/dashboard.ts`; `alert metrics --aggregate` takes the narrower ones metric alerts accept (`METRIC_ALERT_FIELD_COMPLETION`). A word starting with `-` completes flag names, so a descending sort's `-` is typed after completing |
 | Issue as a flag value (`issue merge --into`) | `parseIssueReference` | what `ISSUE_TARGET` completes |
 | Git ref, commit, or repository (`--base-ref`, `--head-sha`, `--repo`, `--commit REPO@SHA`, `set-commits --from`) | `parseVcsRef` (branch and tag names as the remote knows them), `parseGitRef` (refs git resolves locally), `parseGitCommit`, `parseGitRepo` (`owner/repo` of the remotes), `parseRepoCommitList` | read from git in the shell's directory; nothing outside a work tree |
 | A structured value with its own grammar (`issue archive --until`, `api <endpoint>`) | `parseArchiveUntil`, `parseApiEndpoint` | the grammar: `forever`, `auto`, `10x`, `10x/5m`, …; API schema paths one segment at a time, with cached orgs and projects for their placeholders |
diff --git a/packages/cli/src/commands/alert/mutation-utils.ts b/packages/cli/src/commands/alert/mutation-utils.ts
index 059522dd7..d5c0cb7b6 100644
--- a/packages/cli/src/commands/alert/mutation-utils.ts
+++ b/packages/cli/src/commands/alert/mutation-utils.ts
@@ -2,10 +2,9 @@
  * Shared validation and parsing helpers for alert mutation commands (create/edit).
  */
 
-import {
-  DATASET_FIELDS,
-  type DatasetCompletion,
-  type DatasetFields,
+import type {
+  DatasetCompletion,
+  DatasetFields,
 } from "../../lib/completion-target.js";
 import { ValidationError } from "../../lib/errors.js";
 import {
@@ -20,18 +19,78 @@ export const ISSUE_MATCH_MODES = ["all", "any"] as const;
 /** Deprecated dataset names that now route to `spans` with `is_transaction:true`. */
 const TRANSACTION_DATASET_NAMES = new Set(["transaction", "transactions"]);
 
+/*
+ * Aggregates metric alerts accept, from getsentry/sentry (`@sentry/api`
+ * types detector `data_sources` as `Array<unknown>`, so it has none):
+ * - Spans and logs (EAP) skip column checks for the functions in
+ *   `EAP_FUNCTIONS` (`sentry.incidents.logic`); the subscription then
+ *   resolves them with the dataset's EAP definitions, so logs, which have
+ *   no formulas (`sentry.search.eap.ourlogs.aggregates`), take only the
+ *   plain aggregates.
+ * - Errors take a function of no column or of `user`, `release`, `dist`
+ *   (`SUPPORTED_COLUMNS`, `TRANSLATABLE_COLUMNS` in the same module).
+ * - Crash rate takes only `percentage(sessions_crashed|users_crashed,
+ *   sessions|users)` (`CRASH_RATE_ALERT_AGGREGATE_RE` in
+ *   `sentry.snuba.entity_subscription`).
+ */
+
+/** `EAP_FUNCTIONS` that logs also define. */
+const LOG_ALERT_FUNCTIONS = [
+  "count",
+  "count_unique",
+  "avg",
+  "p50",
+  "p75",
+  "p90",
+  "p95",
+  "p99",
+  "max",
+  "min",
+  "sum",
+] as const;
+
+/** `EAP_FUNCTIONS`, all of which spans define. */
+const SPAN_ALERT_FUNCTIONS = [
+  ...LOG_ALERT_FUNCTIONS,
+  "p100",
+  "epm",
+  "eps",
+  "failure_count",
+  "failure_rate",
+  "apdex",
+  "user_misery",
+] as const;
+
+const ERROR_ALERT_FIELDS: DatasetFields = {
+  aggregates: [
+    "count()",
+    "count_unique(user)",
+    "count_unique(release)",
+    "count_unique(dist)",
+  ],
+};
+
+const CRASH_RATE_ALERT_FIELDS: DatasetFields = {
+  aggregates: [
+    "percentage(sessions_crashed,sessions)",
+    "percentage(users_crashed,users)",
+  ],
+};
+
 /**
- * What `--aggregate` completes, by the dataset `--dataset` names. Crash
- * rate (`metrics`, `sessions`) aggregates are not fields of a dataset, so
- * nothing completes for them; without `--dataset`, every dataset's.
+ * What `--aggregate` completes, by the dataset `--dataset` names: the
+ * aggregates metric alerts accept on it, narrower than a widget's. Without
+ * `--dataset`, every dataset's.
  */
 export const METRIC_ALERT_FIELD_COMPLETION: DatasetCompletion = {
   datasets: {
-    errors: DATASET_FIELDS.errors,
-    events: DATASET_FIELDS.errors,
-    spans: DATASET_FIELDS.spans,
-    logs: DATASET_FIELDS.logs,
-  } satisfies Partial<Record<MetricAlertDataset, DatasetFields>>,
+    errors: ERROR_ALERT_FIELDS,
+    events: ERROR_ALERT_FIELDS,
+    spans: { entity: "span-attribute", functions: SPAN_ALERT_FUNCTIONS },
+    logs: { entity: "log-attribute", functions: LOG_ALERT_FUNCTIONS },
+    metrics: CRASH_RATE_ALERT_FIELDS,
+    sessions: CRASH_RATE_ALERT_FIELDS,
+  } satisfies Record<MetricAlertDataset, DatasetFields>,
   aliases: {
     ...METRIC_ALERT_DATASET_ALIASES,
     ...Object.fromEntries(
diff --git a/packages/cli/src/lib/complete.ts b/packages/cli/src/lib/complete.ts
index ef7c09fbb..830832afb 100644
--- a/packages/cli/src/lib/complete.ts
+++ b/packages/cli/src/lib/complete.ts
@@ -45,7 +45,7 @@ import { silenceReadonlyHandling } from "./db/resilient.js";
 import { getEnv } from "./env.js";
 import type { GitCompletion } from "./git.js";
 import { ValidationError } from "./errors.js";
-import { fuzzyMatch } from "./fuzzy.js";
+import { fuzzyMatch, fuzzyMatchSegments } from "./fuzzy.js";
 import type { CommandCompletions } from "./introspect.js";
 import { COMMON_PLATFORMS, VALID_PLATFORMS } from "./platforms.js";
 
@@ -940,9 +940,11 @@ function datasetFieldsOf(slot: CompleterContext): DatasetFields[] {
 }
 
 /**
- * The fields of `datasets` that match the partial: cached entities of the
- * org the command's target names (else of every org), then fixed fields,
- * each once.
+ * The fields of `datasets` that match the partial, one `.` segment at a
+ * time ({@link fuzzyMatchSegments}: `span.d` finds `span.duration`, never
+ * `span.op`): cached entities of the org the command's target names (else
+ * of every org), most recently seen first when nothing is typed, then
+ * fixed fields, each once.
  */
 async function matchFields(
   partial: string,
@@ -950,32 +952,48 @@ async function matchFields(
   datasets: readonly DatasetFields[],
 ): Promise<Completion[]> {
   const found = new Map<string, string | undefined>();
+  const add = (value: string, description?: string) => {
+    if (!found.has(value)) {
+      found.set(value, description);
+    }
+  };
   const orgs = datasets.some(({ entity }) => entity)
     ? await flagOrgs(slot.positionals)
     : [];
   for (const { entity, fields = [] } of datasets) {
-    const cached = entity ? matchCachedEntities(entity, partial, orgs) : [];
-    for (const { value, label } of cached) {
-      if (!found.has(value)) {
-        found.set(value, label);
+    const labels = new Map<string, string | undefined>();
+    for (const { value, label } of entity
+      ? getCachedCompletionEntities(entity, orgs)
+      : []) {
+      if (!labels.has(value)) {
+        labels.set(value, label);
       }
     }
-    for (const value of fuzzyMatch(partial, fields, { mode: "search" })) {
-      if (!found.has(value)) {
-        found.set(value, undefined);
-      }
+    const cached =
+      partial === ""
+        ? [...labels.keys()]
+        : fuzzyMatchSegments(partial, [...labels.keys()]);
+    for (const value of cached.slice(0, MAX_ENTITY_COMPLETIONS)) {
+      add(value, labels.get(value));
+    }
+    for (const value of fuzzyMatchSegments(partial, fields)) {
+      add(value);
     }
   }
   return [...found].map(([value, description]) => ({ value, description }));
 }
 
-/** The aggregate function names of `datasets`, each once. */
-function functionsOf(datasets: readonly DatasetFields[]): string[] {
-  return [...new Set(datasets.flatMap(({ functions = [] }) => functions))];
+/** The aggregate function names, or whole aggregates, of `datasets`, each once. */
+function unionOf(
+  datasets: readonly DatasetFields[],
+  key: "functions" | "aggregates",
+): string[] {
+  return [...new Set(datasets.flatMap((dataset) => dataset[key] ?? []))];
 }
 
 /**
- * Complete an aggregate call. Before the `(`, the function names that
+ * Complete an aggregate call: first the whole aggregates of the dataset
+ * that start with the partial. Before the `(`, the function names that
  * start with the partial, as `fn()`; a name typed in full also completes
  * `fn(field)` for each field, so the shell inserts `fn(` and lists them.
  * After the `(`, `fn(field)` for the fields that match.
@@ -989,7 +1007,10 @@ async function completeAggregates(
   options: { fields: boolean },
 ): Promise<Completion[]> {
   const datasets = datasetFieldsOf(slot);
-  const functions = functionsOf(datasets);
+  const functions = unionOf(datasets, "functions");
+  const fixed = unionOf(datasets, "aggregates")
+    .filter((value) => value.startsWith(partial))
+    .map((value) => ({ value }));
   const calls = async (name: string, argument: string) => [
     ...(argument === "" ? [{ value: `${name}()` }] : []),
     ...(await matchFields(argument, slot, datasets)).map(
@@ -999,13 +1020,17 @@ async function completeAggregates(
   const open = partial.indexOf("(");
   if (open !== -1) {
     const name = partial.slice(0, open);
-    return functions.includes(name)
-      ? await calls(name, partial.slice(open + 1))
-      : [];
+    return [
+      ...fixed,
+      ...(functions.includes(name)
+        ? await calls(name, partial.slice(open + 1))
+        : []),
+    ];
   }
   const names = functions.filter((name) => name.startsWith(partial));
   return [
     ...(options.fields ? await matchFields(partial, slot, datasets) : []),
+    ...fixed,
     ...(names.includes(partial) ? await calls(partial, "") : []),
     ...names
       .filter((name) => name !== partial)
@@ -1024,7 +1049,7 @@ async function completeShorthandAggregates(
   slot: CompleterContext,
 ): Promise<Completion[]> {
   const datasets = datasetFieldsOf(slot);
-  const functions = functionsOf(datasets);
+  const functions = unionOf(datasets, "functions");
   const colon = partial.indexOf(":");
   if (colon === -1) {
     return functions
diff --git a/packages/cli/src/lib/completion-target.ts b/packages/cli/src/lib/completion-target.ts
index 4ae387571..562699430 100644
--- a/packages/cli/src/lib/completion-target.ts
+++ b/packages/cli/src/lib/completion-target.ts
@@ -195,6 +195,8 @@ export type DatasetFields = {
   fields?: readonly string[];
   /** Its aggregate function names */
   functions?: readonly string[];
+  /** Whole aggregates it takes that are not a function of one field */
+  aggregates?: readonly string[];
 };
 
 /**
diff --git a/packages/cli/src/lib/fuzzy.ts b/packages/cli/src/lib/fuzzy.ts
index 6041fc0b6..65a8f9eaa 100644
--- a/packages/cli/src/lib/fuzzy.ts
+++ b/packages/cli/src/lib/fuzzy.ts
@@ -79,6 +79,39 @@ export type FuzzyMatchOptions = {
   mode?: "search";
 };
 
+/**
+ * Score one candidate as {@link fuzzyMatch} ranks it: lower is better.
+ *
+ * @returns The score, or `undefined` when the candidate does not match
+ */
+function matchScore(
+  partial: string,
+  candidate: string,
+  search: boolean,
+): number | undefined {
+  const lowerPartial = partial.toLowerCase();
+  const lowerCandidate = candidate.toLowerCase();
+  if (lowerCandidate === lowerPartial) {
+    return EXACT_SCORE;
+  }
+  if (lowerCandidate.startsWith(lowerPartial)) {
+    return PREFIX_SCORE;
+  }
+  if (lowerCandidate.includes(lowerPartial)) {
+    return CONTAINS_SCORE;
+  }
+  if (search && partial.length < 3) {
+    return;
+  }
+  const dist = editDistance(
+    lowerPartial,
+    lowerCandidate,
+    search ? Math.floor(partial.length / 3) : undefined,
+  );
+  const threshold = Math.max(2, Math.floor(partial.length / 3));
+  return dist <= threshold ? FUZZY_BASE_SCORE + dist : undefined;
+}
+
 /**
  * Rank candidates by relevance to a partial input string.
  *
@@ -115,29 +148,12 @@ export function fuzzyMatch(
     return opts?.maxResults ? sorted.slice(0, opts.maxResults) : sorted;
   }
 
-  const lowerPartial = partial.toLowerCase();
-  const threshold = Math.max(2, Math.floor(partial.length / 3));
   const search = opts?.mode === "search";
   const scored: { candidate: string; score: number }[] = [];
-
   for (const candidate of candidates) {
-    const lowerCandidate = candidate.toLowerCase();
-
-    if (lowerCandidate === lowerPartial) {
-      scored.push({ candidate, score: EXACT_SCORE });
-    } else if (lowerCandidate.startsWith(lowerPartial)) {
-      scored.push({ candidate, score: PREFIX_SCORE });
-    } else if (lowerCandidate.includes(lowerPartial)) {
-      scored.push({ candidate, score: CONTAINS_SCORE });
-    } else if (!search || partial.length >= 3) {
-      const dist = editDistance(
-        lowerPartial,
-        lowerCandidate,
-        search ? Math.floor(partial.length / 3) : undefined,
-      );
-      if (dist <= threshold) {
-        scored.push({ candidate, score: FUZZY_BASE_SCORE + dist });
-      }
+    const score = matchScore(partial, candidate, search);
+    if (score !== undefined) {
+      scored.push({ candidate, score });
     }
   }
 
@@ -148,3 +164,55 @@ export function fuzzyMatch(
   const results = scored.map((s) => s.candidate);
   return opts?.maxResults ? results.slice(0, opts.maxResults) : results;
 }
+
+/**
+ * Rank candidates made of `.`-joined segments (`span.duration`)
+ * by an unfinished input, one segment at a time: the input's Nth segment
+ * is matched against the candidate's Nth, as {@link fuzzyMatch} matches in
+ * search mode. A candidate needs a segment for each one typed, and can
+ * have more after them, so `span.d` finds `span.duration` and
+ * `span.description` but never `span.op`, and `span` finds every `span.*`.
+ *
+ * When `partial` is empty, all candidates are returned (sorted
+ * alphabetically).
+ *
+ * @param partial - The partial input to match against
+ * @param candidates - The list of candidate strings
+ * @returns Candidates sorted by their summed segment scores, then
+ *   alphabetically
+ */
+export function fuzzyMatchSegments(
+  partial: string,
+  candidates: readonly string[],
+): string[] {
+  if (partial === "") {
+    return [...candidates].sort();
+  }
+  const typed = partial.split(".");
+  const scored: { candidate: string; score: number }[] = [];
+  for (const candidate of candidates) {
+    const segments = candidate.split(".");
+    if (segments.length < typed.length) {
+      continue;
+    }
+    let score = 0;
+    for (const [index, segment] of typed.entries()) {
+      const segmentScore =
+        segment === ""
+          ? PREFIX_SCORE
+          : matchScore(segment, segments[index] ?? "", true);
+      if (segmentScore === undefined) {
+        score = Number.POSITIVE_INFINITY;
+        break;
+      }
+      score += segmentScore;
+    }
+    if (Number.isFinite(score)) {
+      scored.push({ candidate, score });
+    }
+  }
+  scored.sort(
+    (a, b) => a.score - b.score || a.candidate.localeCompare(b.candidate),
+  );
+  return scored.map((s) => s.candidate);
+}
diff --git a/packages/cli/test/lib/complete.test.ts b/packages/cli/test/lib/complete.test.ts
index 2dc474394..0345d5af4 100644
--- a/packages/cli/test/lib/complete.test.ts
+++ b/packages/cli/test/lib/complete.test.ts
@@ -1318,7 +1318,27 @@ describe("getCompletions: fields and aggregates of the --dataset given", () => {
     );
   });
 
-  test("alert --aggregate completes calls, not bare fields", async () => {
+  test("inside fn(, fields match one segment at a time", async () => {
+    await seedOrgs([{ slug: "acme", name: "Acme" }]);
+    cacheCompletionEntities("span-attribute", "acme", [
+      { value: "span.op", label: "string" },
+      { value: "span.duration", label: "number" },
+    ]);
+    const field = ["explore", "acme/", "-d", "spans", "-F"];
+
+    // The typed segment never matches past the candidate's: `d` is not `op`
+    expect(await values(field, "avg(span.d")).toEqual(["avg(span.duration)"]);
+    // A typo within the segment being typed still matches
+    expect(await values(field, "avg(span.duraton")).toEqual([
+      "avg(span.duration)",
+    ]);
+    expect((await values(field, "avg(span.")).sort()).toEqual([
+      "avg(span.duration)",
+      "avg(span.op)",
+    ]);
+  });
+
+  test("alert --aggregate completes the calls metric alerts accept", async () => {
     await seedFields();
     const create = ["alert", "metrics", "create", "acme/", "--dataset", "span"];
 
@@ -1328,16 +1348,50 @@ describe("getCompletions: fields and aggregates of the --dataset given", () => {
       "p95()",
       "p99()",
     ]);
+    expect(await values([...create, "--aggregate"], "fail")).toEqual([
+      "failure_count()",
+      "failure_rate()",
+    ]);
     expect(await values([...create, "--aggregate"], "avg(span.d")).toEqual([
       "avg(span.duration)",
     ]);
-    // Crash rate aggregates are not fields of a dataset
-    expect(
-      await values(
-        ["alert", "metrics", "create", "--dataset", "sessions", "--aggregate"],
-        "",
-      ),
-    ).toEqual([]);
+    // Not an alert function, though widgets take it
+    expect(await values([...create, "--aggregate"], "any")).toEqual([]);
+  });
+
+  test("alert --aggregate narrows to each dataset's aggregates", async () => {
+    await seedFields();
+    const aggregate = (dataset: string) => [
+      "alert",
+      "metrics",
+      "create",
+      "acme/",
+      "--dataset",
+      dataset,
+      "--aggregate",
+    ];
+
+    // Logs have no formulas: no epm, failure_rate, or p100
+    expect(await values(aggregate("logs"), "p10")).toEqual([]);
+    expect(await values(aggregate("log"), "e")).toEqual([]);
+    // Errors take no column but user, release, dist, not their tag keys
+    expect(await values(aggregate("errors"), "")).toEqual([
+      "count()",
+      "count_unique(user)",
+      "count_unique(release)",
+      "count_unique(dist)",
+    ]);
+    expect(await values(aggregate("errors"), "count_unique(u")).toEqual([
+      "count_unique(user)",
+    ]);
+    // Crash rate takes only crash-free percentages
+    expect(await values(aggregate("sessions"), "")).toEqual([
+      "percentage(sessions_crashed,sessions)",
+      "percentage(users_crashed,users)",
+    ]);
+    expect(await values(aggregate("metrics"), "percentage(u")).toEqual([
+      "percentage(users_crashed,users)",
+    ]);
   });
 
   test("dashboard widgets complete shorthand aggregates and group-by columns", async () => {
diff --git a/packages/cli/test/lib/fuzzy.test.ts b/packages/cli/test/lib/fuzzy.test.ts
index 3513d603f..b562ca1d0 100644
--- a/packages/cli/test/lib/fuzzy.test.ts
+++ b/packages/cli/test/lib/fuzzy.test.ts
@@ -8,7 +8,11 @@
 
 import { array, assert as fcAssert, property, string } from "fast-check";
 import { describe, expect, test } from "vitest";
-import { fuzzyMatch, levenshtein } from "../../src/lib/fuzzy.js";
+import {
+  fuzzyMatch,
+  fuzzyMatchSegments,
+  levenshtein,
+} from "../../src/lib/fuzzy.js";
 import { DEFAULT_NUM_RUNS } from "../model-based/helpers.js";
 
 // Short strings to keep Levenshtein computation fast in property tests
@@ -190,6 +194,56 @@ describe("fuzzyMatch: search mode", () => {
   });
 });
 
+describe("fuzzyMatchSegments", () => {
+  const fields = [
+    "span.op",
+    "span.duration",
+    "span.description",
+    "http.method",
+    "span",
+  ];
+
+  test("matches the typed segment within the candidate's, never past it", () => {
+    expect(fuzzyMatchSegments("span.d", fields)).toEqual([
+      "span.description",
+      "span.duration",
+    ]);
+    expect(fuzzyMatchSegments("span.d", ["span.op"])).toEqual([]);
+  });
+
+  test("tolerates a typo within a segment", () => {
+    expect(fuzzyMatchSegments("span.duraton", fields)).toEqual([
+      "span.duration",
+    ]);
+    expect(fuzzyMatchSegments("spn.duration", fields)).toEqual([
+      "span.duration",
+    ]);
+  });
+
+  test("a segment typed in full matches the candidates under it", () => {
+    expect(fuzzyMatchSegments("span", fields)).toEqual([
+      "span",
+      "span.description",
+      "span.duration",
+      "span.op",
+    ]);
+    expect(fuzzyMatchSegments("span.", fields)).toEqual([
+      "span.description",
+      "span.duration",
+      "span.op",
+    ]);
+  });
+
+  test("needs a candidate segment for each typed one", () => {
+    expect(fuzzyMatchSegments("span.op.x", fields)).toEqual([]);
+    expect(fuzzyMatchSegments("duration", fields)).toEqual([]);
+  });
+
+  test("returns every candidate, sorted, for an empty partial", () => {
+    expect(fuzzyMatchSegments("", ["b.a", "a.b"])).toEqual(["a.b", "b.a"]);
+  });
+});
+
 describe("property: fuzzyMatch", () => {
   test("exact candidate is always in the result", () => {
     const candidatesArb = array(shortStringArb, {