ktx/packages/cli/src/context/scan/relationship-benchmark-report.ts
Andrey Avtomonov 2366b00301
chore(workspace): gate dead-code with knip production mode (#196)
* refactor(workspace): relocate @ktx/llm source into packages/cli/src/llm

* refactor(workspace): rewrite @ktx/llm imports to relative paths

* refactor(workspace): fold internal packages into cli

* chore(workspace): gate dead-code with knip production mode

Turn on production-mode knip plus an autofix run in pre-commit and the
`pnpm dead-code` script, document the `/** @internal */` convention for
test-only exports in AGENTS.md, annotate test-only exports across the
CLI with that JSDoc, and drop dead exports/wrappers the new gate
surfaced (e.g. `cli-project.ts`, `lookerRuntimeSourceToFileAdapterSource`,
`createLocalScanEnrichmentProvidersFromConfig`,
`PGLITE_OWNER_PROCESS_BACKEND_CAPABILITIES`, stale type re-exports).
Replace the loose `ignoreIssues` allowlist in `knip.json` with explicit
production entries so cross-package barrel leaks are caught.

* refactor(cli): delete internal barrel index.ts files

The 34 `index.ts` re-export barrels inside `packages/cli/src/` were
holdovers from the pre-fold multi-workspace structure. Post-fold-in they
served no production purpose: external consumers go through the single
package main entry, and in-repo callers mostly imported through them
only because the path was short. Internally, knip flagged most barrel
re-exports as production-dead (only reached via tests).

This change:
- Deletes every internal barrel except `packages/cli/src/index.ts`
  (the published package entry).
- Rewrites ~270 source/test files to import each name directly from
  the file that defines it.
- Moves `tools/warehouse-verification/index.ts` to
  `create-warehouse-verification-tools.ts` (the function it defined
  locally) and updates its single consumer.
- Renames `search/backend-conformance.ts` → `.test-utils.ts` to match
  the existing test-helper file convention.
- Deletes 13 dead test-only chains (dbt-descriptions/*,
  live-database/extracted-schema, live-database/structural-sync,
  relationship-* feedback/review chain) plus their tests and a
  cascading orphan integration test.
- Updates test mocks that pointed at deleted barrel paths
  (notion-client, connector barrels in scan/local-scan-connectors
  tests) to mock the source files instead.
- Points the maintainer benchmark script
  (`scripts/relationship-benchmark-report.mjs`) at source files
  instead of `dist/context/scan/index.js`.
- Drops the barrel `!` entries from `knip.json`; adds explicit
  production entries only for the benchmark code reached via dist by
  the maintainer script.

Net: 413 files changed, ~1.2k insertions, ~9.4k deletions.

`pnpm run dead-code` (Biome + knip default + knip production) and
`pnpm run type-check` are clean; 2277 tests pass.

* refactor(workspace): rename @ktx/cli to @kaelio/ktx and pack it directly

Promote the CLI workspace package to the public name `@kaelio/ktx` and
drop the separate `scripts/build-public-npm-package.mjs` wrapper. The
CLI package is now publishable in place (`publishConfig.access: public`,
`provenance: true`), so artifact packing uses `pnpm pack` against
`packages/cli/` instead of assembling a parallel package tree.

Updates all workspace filter invocations, docs, tests, and release
readiness checks to reference the new package name, and folds the
tarball-name helper into `scripts/public-npm-release-metadata.mjs`.

* docs: align "agent clients" and "data agents" terminology

Replace "client agents" with "agent clients" and "database agents" with
"data agents" across AGENTS.md, README.md, the docs-site copy, and the
matching setup-agents test description, matching the canonical
vocabulary in docs/terminology.md.

Also moves packages/cli/tsconfig.json's tsBuildInfoFile from
node_modules/.cache/ to dist/.tsbuildinfo so incremental builds survive
node_modules reinstalls.

* refactor(release): single source of truth for package version

Make packages/cli/package.json the single source of truth for the
@kaelio/ktx version. publicNpmPackageVersion() now reads it directly,
so artifact filenames, release-readiness checks, and the Python wheel
version all derive from one field. The duplicate
release-policy.json.publicNpmPackageVersion is removed.

Previously the two fields could drift: tarballs were named
kaelio-ktx-0.4.1.tgz while internally containing
@kaelio/ktx@0.0.0-private.

- update-public-release-version.mjs rewrites both Python pyproject.toml
  files (ktx-daemon, ktx-sl) alongside the npm package.jsons,
  normalizing the version for PEP 440 (e.g. 0.1.0-rc.2 -> 0.1.0rc2).
- semantic-release-config.cjs adds the two pyproject.toml files to
  @semantic-release/git assets so the release commit back to main
  carries every version source in lockstep.
- The six "?? '0.0.0-private'" fallback literals across the CLI are
  replaced with "?? getKtxCliPackageInfo().version", and
  createDefaultKtxMcpServer makes its version arg required.
- docs/release.md describes the actual commit-back model: the dev tree
  always reflects the most recent release; no sentinel pin to
  maintain.

Verified: pnpm run artifacts:build now produces
kaelio-ktx-0.4.1.tgz and kaelio_ktx-0.4.1-py3-none-any.whl with
@kaelio/ktx@0.4.1 inside. Full type-check, dead-code, and
2287 vitests + 173 script tests pass.

* refactor(cli): inject embedding provider resolution and detect sentence-transformers runtime

Make resolveProjectEmbeddingProvider and runtimeIo injectable in ingest and
scan command entrypoints so tests can stub them, and teach
resolvePublicIngestRuntimeRequirements to flag the local-embeddings runtime
feature when ktx.yaml selects sentence-transformers.

* chore(cli): mark buildLocalStatsStatus and LocalStatsStatus as @internal

Both symbols are consumed only by status-project.test.ts. Annotating with
/** @internal */ keeps knip's production-mode check clean without changing
runtime behavior.

* fix(cli): use real package metadata in print-command-tree

The stubbed package name embedded a forbidden product identifier that
tripped the boundary check in CI. Read the metadata from package.json
instead — keeps the rendered tree unchanged and removes a duplicate
source of truth.

* feat(cli): show embedding coverage in `ktx status`, drop duplicate disk counts

Inline `(N embedded)` next to the Wiki scope counts and Semantic-layer
source counts, computed with `SUM(embedding_json IS NOT NULL)` over
`knowledge_pages` and `local_sl_sources`. Rename the "Knowledge" label to
"Wiki" (canonical per `docs/terminology.md`) and rename the matching
`localStats.knowledgePages` field to `localStats.wikiPages`.

Drop `wiki=N md` and `semantic-layer=N yaml` from the Disk row — those
duplicated the per-surface rows above. Disk now reports only actual byte
usage (db, cache, raw-sources). The unused `wikiGlobalMarkdownCount` /
`semanticLayerYamlCount` fields, the `isMarkdownEntry` / `isYamlEntry`
helpers, and the `filter` arg on `summarizeDir` are removed.
2026-05-21 15:28:58 +02:00

363 lines
12 KiB
TypeScript

import { isKtxRelationshipBenchmarkTuningEligible } from './relationship-benchmarks.js';
import type {
KtxRelationshipBenchmarkCaseResult,
KtxRelationshipBenchmarkFixture,
KtxRelationshipBenchmarkMode,
KtxRelationshipBenchmarkSuiteResult,
} from './relationship-benchmarks.js';
export type KtxRelationshipBenchmarkReportCaseStatus = 'run' | 'validation_blocked' | 'not_run';
export interface KtxRelationshipBenchmarkReportCase {
fixtureId: string;
fixtureName: string;
tier: string;
mode: KtxRelationshipBenchmarkMode;
status: KtxRelationshipBenchmarkReportCaseStatus;
reason: string | null;
tuningEligible: boolean;
metrics: {
pkRecall: number | null;
fkRecall: number | null;
acceptedOrReviewRecall: number | null;
acceptedFalsePositiveCount: number | null;
sqlQueries: number | null;
llmCalls: number | null;
runtimeSeconds: number | null;
};
falsePositives: {
pk: string[];
fk: string[];
};
falseNegatives: {
pk: string[];
fk: string[];
};
skippedComposite: {
pk: string[];
fk: string[];
};
}
export interface KtxRelationshipBenchmarkReport {
generatedAt: string;
headline: {
caseCount: number;
headlineCaseCount: number;
headlinePkRecall: number;
headlineFkRecall: number;
headlineAcceptedOrReviewRecall: number;
acceptedFalsePositiveCount: number;
validationBlockedCount: number;
};
cases: KtxRelationshipBenchmarkReportCase[];
}
function key(fixtureId: string, mode: KtxRelationshipBenchmarkMode): string {
return `${fixtureId}:${mode}`;
}
function fixed(value: number | null): string {
return value === null ? '-' : value.toFixed(3);
}
function reportCaseReason(input: {
fixture: KtxRelationshipBenchmarkFixture;
result: KtxRelationshipBenchmarkCaseResult;
}): string | null {
if (input.result.validationBlocked) {
return 'validation unavailable for this benchmark mode';
}
if (input.fixture.validationBudget !== undefined && input.result.predicted.reviewFk.length > 0) {
return `review candidate validation reasons: validation_unattempted (${input.result.predicted.reviewFk.length})`;
}
return null;
}
function reportCaseFromResult(input: {
fixture: KtxRelationshipBenchmarkFixture;
mode: KtxRelationshipBenchmarkMode;
result: KtxRelationshipBenchmarkCaseResult;
}): KtxRelationshipBenchmarkReportCase {
const status = input.result.validationBlocked ? 'validation_blocked' : 'run';
return {
fixtureId: input.fixture.id,
fixtureName: input.fixture.name,
tier: input.fixture.tier,
mode: input.mode,
status,
reason: reportCaseReason({ fixture: input.fixture, result: input.result }),
tuningEligible: isKtxRelationshipBenchmarkTuningEligible({
fixture: input.fixture,
mode: input.mode,
validationBlocked: input.result.validationBlocked,
}),
metrics: {
pkRecall: input.result.metrics.pkRecall,
fkRecall: input.result.metrics.fkRecall,
acceptedOrReviewRecall: input.result.metrics.acceptedOrReviewRecall,
acceptedFalsePositiveCount: input.result.metrics.acceptedFalsePositiveCount,
sqlQueries: input.result.metrics.sqlQueries,
llmCalls: input.result.metrics.llmCalls,
runtimeSeconds: input.result.metrics.runtimeSeconds,
},
falsePositives: input.result.falsePositives,
falseNegatives: input.result.falseNegatives,
skippedComposite: input.result.skippedComposite,
};
}
function notRunCase(input: {
fixture: KtxRelationshipBenchmarkFixture;
mode: KtxRelationshipBenchmarkMode;
reason: string;
}): KtxRelationshipBenchmarkReportCase {
return {
fixtureId: input.fixture.id,
fixtureName: input.fixture.name,
tier: input.fixture.tier,
mode: input.mode,
status: 'not_run',
reason: input.reason,
tuningEligible: false,
metrics: {
pkRecall: null,
fkRecall: null,
acceptedOrReviewRecall: null,
acceptedFalsePositiveCount: null,
sqlQueries: null,
llmCalls: null,
runtimeSeconds: null,
},
falsePositives: { pk: [], fk: [] },
falseNegatives: { pk: [], fk: [] },
skippedComposite: { pk: [], fk: [] },
};
}
export function buildKtxRelationshipBenchmarkReport(input: {
fixtures: readonly KtxRelationshipBenchmarkFixture[];
suite: KtxRelationshipBenchmarkSuiteResult;
modes: readonly KtxRelationshipBenchmarkMode[];
generatedAt?: string;
}): KtxRelationshipBenchmarkReport {
const resultsByKey = new Map(input.suite.cases.map((result) => [key(result.fixtureId, result.mode), result]));
const cases: KtxRelationshipBenchmarkReportCase[] = [];
for (const fixture of input.fixtures) {
const selectedModes = new Set(fixture.defaultModes);
for (const mode of input.modes) {
const result = resultsByKey.get(key(fixture.id, mode));
if (result) {
cases.push(reportCaseFromResult({ fixture, mode, result }));
continue;
}
cases.push(
notRunCase({
fixture,
mode,
reason: selectedModes.has(mode) ? 'mode produced no benchmark result' : 'mode not selected by fixture defaultModes',
}),
);
}
}
return {
generatedAt: input.generatedAt ?? new Date().toISOString(),
headline: {
caseCount: input.suite.aggregate.caseCount,
headlineCaseCount: input.suite.aggregate.headlineCaseCount,
headlinePkRecall: input.suite.aggregate.headlinePkRecall,
headlineFkRecall: input.suite.aggregate.headlineFkRecall,
headlineAcceptedOrReviewRecall: input.suite.aggregate.headlineAcceptedOrReviewRecall,
acceptedFalsePositiveCount: input.suite.cases.reduce(
(sum, result) => sum + result.metrics.acceptedFalsePositiveCount,
0,
),
validationBlockedCount: input.suite.validationBlockedCases.length,
},
cases,
};
}
type KtxRelationshipBenchmarkFailureSelector = (
item: KtxRelationshipBenchmarkReportCase,
) => readonly string[];
function sortedFailureLines(input: {
cases: readonly KtxRelationshipBenchmarkReportCase[];
select: KtxRelationshipBenchmarkFailureSelector;
}): string[] {
return input.cases
.flatMap((item) =>
input.select(item).map((value) => ({
fixtureId: item.fixtureId,
mode: item.mode,
status: item.status,
value,
})),
)
.sort((left, right) => {
const leftKey = `${left.fixtureId}:${left.mode}:${left.status}:${left.value}`;
const rightKey = `${right.fixtureId}:${right.mode}:${right.status}:${right.value}`;
return leftKey.localeCompare(rightKey);
})
.map((item) => `- \`${item.fixtureId}\` / \`${item.mode}\` / \`${item.status}\`: ${item.value}`);
}
function failureBlock(input: {
title: string;
cases: readonly KtxRelationshipBenchmarkReportCase[];
select: KtxRelationshipBenchmarkFailureSelector;
}): string[] {
const values = sortedFailureLines({ cases: input.cases, select: input.select });
return ['', `### ${input.title}`, '', ...(values.length > 0 ? values : ['- none'])];
}
function headlineFailureContextBlocks(report: KtxRelationshipBenchmarkReport): string[] {
const headlineCases = report.cases.filter((item) => item.tuningEligible);
const remainingPkMisses = sortedFailureLines({
cases: headlineCases,
select: (item) => item.falseNegatives.pk,
});
const remainingFkMisses = sortedFailureLines({
cases: headlineCases,
select: (item) => item.falseNegatives.fk,
});
return [
'',
'## Headline Failure Context',
'',
'Remaining headline misses after this run are listed here so recall gains and still-open algorithmic gaps are visible in the regenerated evidence report.',
'',
`- Remaining headline false-negative PKs: ${remainingPkMisses.length}`,
`- Remaining headline false-negative FKs: ${remainingFkMisses.length}`,
'',
'### Remaining Headline False Negative PKs',
'',
...(remainingPkMisses.length > 0 ? remainingPkMisses : ['- none']),
'',
'### Remaining Headline False Negative FKs',
'',
...(remainingFkMisses.length > 0 ? remainingFkMisses : ['- none']),
];
}
function failureDetailBlocks(report: KtxRelationshipBenchmarkReport): string[] {
const headlineCases = report.cases.filter((item) => item.tuningEligible);
const otherCases = report.cases.filter((item) => !item.tuningEligible);
return [
'',
'## Failure Details',
...failureBlock({
title: 'Headline False Positive PKs',
cases: headlineCases,
select: (item) => item.falsePositives.pk,
}),
...failureBlock({
title: 'Headline False Positive FKs',
cases: headlineCases,
select: (item) => item.falsePositives.fk,
}),
...failureBlock({
title: 'Headline False Negative PKs',
cases: headlineCases,
select: (item) => item.falseNegatives.pk,
}),
...failureBlock({
title: 'Headline False Negative FKs',
cases: headlineCases,
select: (item) => item.falseNegatives.fk,
}),
...failureBlock({
title: 'Other False Positive PKs',
cases: otherCases,
select: (item) => item.falsePositives.pk,
}),
...failureBlock({
title: 'Other False Positive FKs',
cases: otherCases,
select: (item) => item.falsePositives.fk,
}),
...failureBlock({
title: 'Other False Negative PKs',
cases: otherCases,
select: (item) => item.falseNegatives.pk,
}),
...failureBlock({
title: 'Other False Negative FKs',
cases: otherCases,
select: (item) => item.falseNegatives.fk,
}),
];
}
function compositeSkipBlocks(report: KtxRelationshipBenchmarkReport): string[] {
const headlineCases = report.cases.filter((item) => item.tuningEligible);
return [
'',
'## Composite Ground Truth Skips',
...failureBlock({
title: 'Skipped Composite PKs',
cases: headlineCases,
select: (item) => item.skippedComposite.pk,
}),
...failureBlock({
title: 'Skipped Composite FKs',
cases: headlineCases,
select: (item) => item.skippedComposite.fk,
}),
];
}
export function formatKtxRelationshipBenchmarkReportMarkdown(report: KtxRelationshipBenchmarkReport): string {
const lines = [
'# KTX Relationship Discovery Benchmark Evidence',
'',
`Generated: ${report.generatedAt}`,
'',
'## Headline',
'',
`- Cases run: ${report.headline.caseCount}`,
`- Headline cases: ${report.headline.headlineCaseCount}`,
`- Headline PK recall: ${fixed(report.headline.headlinePkRecall)}`,
`- Headline FK recall: ${fixed(report.headline.headlineFkRecall)}`,
`- Headline accepted-or-review recall: ${fixed(report.headline.headlineAcceptedOrReviewRecall)}`,
`- Accepted false positives: ${report.headline.acceptedFalsePositiveCount}`,
`- Validation-blocked cases: ${report.headline.validationBlockedCount}`,
'',
'## Cases',
'',
'| Fixture | Tier | Mode | Status | Tuning Eligible | PK Recall | FK Recall | Accepted+Review Recall | Accepted FP | Reason |',
'| --- | --- | --- | --- | --- | ---: | ---: | ---: | ---: | --- |',
];
for (const item of report.cases) {
lines.push(
[
`| ${item.fixtureId}`,
item.tier,
item.mode,
item.status,
item.tuningEligible ? 'yes' : 'no',
fixed(item.metrics.pkRecall),
fixed(item.metrics.fkRecall),
fixed(item.metrics.acceptedOrReviewRecall),
String(item.metrics.acceptedFalsePositiveCount ?? '-'),
`${item.reason ?? ''} |`,
].join(' | '),
);
}
lines.push(...headlineFailureContextBlocks(report));
lines.push(...failureDetailBlocks(report));
lines.push(...compositeSkipBlocks(report));
lines.push('');
return `${lines.join('\n')}\n`;
}