ktx/packages/cli/src/setup-embeddings.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

495 lines
17 KiB
TypeScript

import { readFile, writeFile } from 'node:fs/promises';
import { resolveKtxConfigReference } from './context/core/config-reference.js';
import { type KtxProjectConfig, type KtxProjectEmbeddingConfig, serializeKtxProjectConfig } from './context/project/config.js';
import { loadKtxProject } from './context/project/project.js';
import { markKtxSetupStateStepComplete, readKtxSetupState } from './context/project/setup-config.js';
import type { KtxEmbeddingConfig } from './llm/types.js';
import { type KtxEmbeddingHealthCheckResult, runKtxEmbeddingHealthCheck } from './llm/embedding-health.js';
import type { KtxCliIo } from './cli-runtime.js';
import { createStaticCliSpinner, type KtxCliSpinner } from './clack.js';
import {
ensureManagedLocalEmbeddingsDaemon,
managedLocalEmbeddingHealthConfig,
type ManagedLocalEmbeddingsDaemon,
} from './managed-local-embeddings.js';
import type { KtxManagedPythonInstallPolicy } from './managed-python-command.js';
import { withTextInputNavigation } from './prompt-navigation.js';
import { envCredentialReference, writeProjectLocalSecretReference } from './setup-secrets.js';
import {
createKtxSetupPromptAdapter,
type KtxSetupPromptOption,
} from './setup-prompts.js';
export type KtxSetupEmbeddingBackend = 'openai' | 'sentence-transformers';
export interface KtxSetupEmbeddingsArgs {
projectDir: string;
inputMode: 'auto' | 'disabled';
cliVersion: string;
runtimeInstallPolicy: KtxManagedPythonInstallPolicy;
embeddingBackend?: KtxSetupEmbeddingBackend;
embeddingApiKeyEnv?: string;
embeddingApiKeyFile?: string;
forcePrompt?: boolean;
showPromptInstructions?: boolean;
skipEmbeddings: boolean;
}
export type KtxSetupEmbeddingsResult =
| { status: 'ready'; projectDir: string }
| { status: 'skipped'; projectDir: string }
| { status: 'back'; projectDir: string }
| { status: 'missing-input'; projectDir: string }
| { status: 'failed'; projectDir: string };
/** @internal */
export interface KtxSetupEmbeddingsPromptAdapter {
select(options: { message: string; options: KtxSetupPromptOption[] }): Promise<string>;
password(options: { message: string }): Promise<string | undefined>;
cancel(message: string): void;
}
export interface KtxSetupEmbeddingsDeps {
env?: NodeJS.ProcessEnv;
prompts?: KtxSetupEmbeddingsPromptAdapter;
healthCheck?: (config: KtxEmbeddingConfig) => Promise<KtxEmbeddingHealthCheckResult>;
ensureLocalEmbeddings?: (options: {
cliVersion: string;
projectDir: string;
installPolicy: KtxManagedPythonInstallPolicy;
io: KtxCliIo;
}) => Promise<ManagedLocalEmbeddingsDaemon>;
spinner?: () => KtxCliSpinner;
}
type BackendChoice = KtxSetupEmbeddingBackend | 'back';
const DEFAULTS: Record<
KtxSetupEmbeddingBackend,
{ model: string; dimensions: number; envName?: string; baseUrl?: string; pathPrefix?: string }
> = {
openai: { model: 'text-embedding-3-small', dimensions: 1536, envName: 'OPENAI_API_KEY' },
'sentence-transformers': {
model: 'all-MiniLM-L6-v2',
dimensions: 384,
baseUrl: 'http://127.0.0.1:8765',
pathPrefix: '',
},
};
const LOCAL_EMBEDDING_BACKEND: KtxSetupEmbeddingBackend = 'sentence-transformers';
const EMBEDDING_OPTION_PROMPT_CONTEXT =
'KTX uses embeddings for semantic search over semantic-layer sources, wiki context, schema metadata, ' +
'and relationship evidence.';
const LOCAL_EMBEDDING_HEALTH_TIMEOUT_MS = 120_000;
const LOCAL_EMBEDDING_STDERR_TAIL_LINES = 40;
function createPromptAdapter(): KtxSetupEmbeddingsPromptAdapter {
return createKtxSetupPromptAdapter({ selectCancelValue: 'back' });
}
async function hasCompletedEmbeddings(projectDir: string, config: KtxProjectConfig): Promise<boolean> {
return (
(await readKtxSetupState(projectDir)).completed_steps.includes('embeddings') &&
config.ingest.embeddings.backend !== 'none' &&
typeof config.ingest.embeddings.model === 'string' &&
config.ingest.embeddings.model.length > 0 &&
config.ingest.embeddings.dimensions > 0
);
}
function buildProjectEmbeddingConfig(input: {
backend: KtxSetupEmbeddingBackend;
model: string;
dimensions: number;
credentialRef?: string;
}): KtxProjectEmbeddingConfig {
if (input.backend === 'openai') {
return {
backend: 'openai',
model: input.model,
dimensions: input.dimensions,
openai: {
...(input.credentialRef ? { api_key: input.credentialRef } : {}),
},
};
}
const defaults = DEFAULTS[input.backend];
return {
backend: input.backend,
model: input.model,
dimensions: input.dimensions,
sentenceTransformers: {
base_url: defaults.baseUrl ?? '',
pathPrefix: defaults.pathPrefix ?? '',
},
};
}
function buildHealthConfig(input: {
backend: KtxSetupEmbeddingBackend;
model: string;
dimensions: number;
credentialValue?: string;
}): KtxEmbeddingConfig {
if (input.backend === 'openai') {
return {
backend: 'openai',
model: input.model,
dimensions: input.dimensions,
openai: {
...(input.credentialValue ? { apiKey: input.credentialValue } : {}),
},
};
}
const defaults = DEFAULTS[input.backend];
return {
backend: input.backend,
model: input.model,
dimensions: input.dimensions,
sentenceTransformers: {
baseURL: defaults.baseUrl ?? '',
pathPrefix: defaults.pathPrefix ?? '',
},
};
}
function embeddingBackendDisplayName(backend: KtxSetupEmbeddingBackend): string {
if (backend === 'openai') {
return 'OpenAI';
}
return 'sentence-transformers';
}
async function persistEmbeddingConfig(projectDir: string, embeddings: KtxProjectEmbeddingConfig): Promise<void> {
const project = await loadKtxProject({ projectDir });
const config = {
...project.config,
ingest: {
...project.config.ingest,
embeddings,
},
scan: {
...project.config.scan,
enrichment: {
...project.config.scan.enrichment,
embeddings,
},
},
};
await writeFile(project.configPath, serializeKtxProjectConfig(config), 'utf-8');
await markKtxSetupStateStepComplete(projectDir, 'embeddings');
}
async function chooseCredentialRef(
backend: Extract<KtxSetupEmbeddingBackend, 'openai'>,
args: KtxSetupEmbeddingsArgs,
io: KtxCliIo,
deps: KtxSetupEmbeddingsDeps,
): Promise<{ status: 'ready'; ref: string; value: string } | { status: 'back' | 'missing-input' }> {
const env = deps.env ?? process.env;
if (args.embeddingApiKeyEnv) {
const ref = envCredentialReference(args.embeddingApiKeyEnv);
const value = resolveKtxConfigReference(ref, env);
if (!value) {
io.stderr.write(`Missing embedding API key: ${args.embeddingApiKeyEnv} is not set.\n`);
return { status: 'missing-input' };
}
return { status: 'ready', ref, value };
}
if (args.embeddingApiKeyFile) {
const ref = `file:${args.embeddingApiKeyFile}`;
let value: string | undefined;
try {
value = resolveKtxConfigReference(ref, env);
} catch {
value = undefined;
}
if (!value) {
io.stderr.write(`Missing embedding API key file: ${args.embeddingApiKeyFile}\n`);
return { status: 'missing-input' };
}
return { status: 'ready', ref, value };
}
if (args.inputMode === 'disabled') {
io.stderr.write('Missing embedding API key: pass --embedding-api-key-env or --embedding-api-key-file.\n');
return { status: 'missing-input' };
}
const defaultEnv = DEFAULTS[backend].envName ?? 'EMBEDDING_API_KEY';
const prompts = deps.prompts ?? createPromptAdapter();
const choice = await prompts.select({
message: `How should KTX find your ${embeddingBackendDisplayName(backend)} embedding API key?`,
options: [
{ value: 'env', label: `Use ${defaultEnv} from the environment` },
{ value: 'paste', label: 'Paste a key and save it as a local secret file' },
{ value: 'back', label: 'Back' },
],
});
if (choice === 'back') {
return { status: 'back' };
}
if (choice === 'paste') {
io.stdout.write(
`${[
`KTX will save the key in .ktx/secrets/${backend}-api-key with local file permissions,`,
'then write a file: reference in ktx.yaml.',
].join(' ')}\n`,
);
const value = await prompts.password({ message: withTextInputNavigation(`${backend} embedding API key`) });
if (value === undefined) {
return { status: 'back' };
}
if (!value.trim()) {
return { status: 'missing-input' };
}
const ref = await writeProjectLocalSecretReference({
projectDir: args.projectDir,
fileName: `${backend}-api-key`,
value,
});
return { status: 'ready', ref, value: value.trim() };
}
const ref = envCredentialReference(defaultEnv);
const value = resolveKtxConfigReference(ref, env);
if (!value) {
io.stderr.write(`Missing embedding API key: ${defaultEnv} is not set.\n`);
return { status: 'missing-input' };
}
return { status: 'ready', ref, value };
}
async function chooseEmbeddingBackend(
args: KtxSetupEmbeddingsArgs,
deps: KtxSetupEmbeddingsDeps,
): Promise<BackendChoice> {
if (args.embeddingBackend) {
return args.embeddingBackend;
}
if (args.inputMode === 'disabled') {
return LOCAL_EMBEDDING_BACKEND;
}
const choice = await (deps.prompts ?? createPromptAdapter()).select({
message: `Which embedding option should KTX use?\n\n${EMBEDDING_OPTION_PROMPT_CONTEXT}`,
options: [
{ value: 'sentence-transformers', label: 'Local sentence-transformers embeddings' },
{ value: 'openai', label: 'OpenAI embeddings', hint: 'recommended' },
{ value: 'back', label: 'Back' },
],
});
if (choice === 'openai' || choice === 'sentence-transformers' || choice === 'back') {
return choice;
}
return 'back';
}
async function readLocalEmbeddingDaemonStderrTail(stderrLog: string | undefined): Promise<string[]> {
if (!stderrLog) {
return [];
}
try {
const lines = (await readFile(stderrLog, 'utf8'))
.split(/\r?\n/)
.map((line) => line.trimEnd())
.filter((line) => line.trim().length > 0);
return lines.slice(-LOCAL_EMBEDDING_STDERR_TAIL_LINES);
} catch {
return [];
}
}
function localEmbeddingSetupMessage(message: string, stderrTail: string[] = []): string {
const lines = [
`Local embedding health check failed: ${message}`,
'Local embeddings use the KTX-managed Python runtime.',
'Prepare the runtime with: ktx admin runtime start --feature local-embeddings',
'Use --yes with setup to install and start the runtime without prompting.',
'The first run may download Python packages and the all-MiniLM-L6-v2 model.',
];
if (stderrTail.length > 0) {
lines.push('Recent KTX daemon stderr:', ...stderrTail);
}
return lines.join('\n');
}
async function promptAfterLocalEmbeddingFailure(
deps: KtxSetupEmbeddingsDeps,
): Promise<'retry' | Extract<KtxSetupEmbeddingBackend, 'openai'> | 'back'> {
const choice = await (deps.prompts ?? createPromptAdapter()).select({
message: 'Local embeddings are not reachable. Start the local KTX daemon, then retry.',
options: [
{ value: 'retry', label: 'Retry' },
{ value: 'openai', label: 'Use OpenAI embeddings' },
{ value: 'back', label: 'Back' },
],
});
if (choice === 'openai' || choice === 'back') {
return choice;
}
return 'retry';
}
function healthCheckStartText(backend: KtxSetupEmbeddingBackend, model: string, dimensions: number): string {
if (backend === LOCAL_EMBEDDING_BACKEND) {
return `Testing local embeddings (${model})`;
}
return `Checking ${backend} embeddings (${model}, ${dimensions} dimensions).`;
}
function startHealthCheckProgress(
spinner: KtxCliSpinner,
message: string,
): { succeed(msg: string): void; fail(msg: string): void } {
spinner.start(message);
return {
succeed(msg: string) {
spinner.stop(msg);
},
fail(msg: string) {
spinner.error(msg);
},
};
}
export async function runKtxSetupEmbeddingsStep(
args: KtxSetupEmbeddingsArgs,
io: KtxCliIo,
deps: KtxSetupEmbeddingsDeps = {},
): Promise<KtxSetupEmbeddingsResult> {
if (args.skipEmbeddings) {
io.stdout.write('│ Embeddings setup skipped.\n');
return { status: 'skipped', projectDir: args.projectDir };
}
const project = await loadKtxProject({ projectDir: args.projectDir });
if (
args.forcePrompt !== true &&
(await hasCompletedEmbeddings(args.projectDir, project.config)) &&
!args.embeddingBackend &&
!args.embeddingApiKeyEnv &&
!args.embeddingApiKeyFile
) {
io.stdout.write(`│ Embeddings ready: yes (${project.config.ingest.embeddings.model})\n`);
return { status: 'ready', projectDir: args.projectDir };
}
const healthCheck =
deps.healthCheck ??
((config: KtxEmbeddingConfig) =>
runKtxEmbeddingHealthCheck(config, { timeoutMs: LOCAL_EMBEDDING_HEALTH_TIMEOUT_MS }));
let selectedBackend: KtxSetupEmbeddingBackend | undefined;
while (true) {
if (!selectedBackend) {
const backend = await chooseEmbeddingBackend(args, deps);
if (backend === 'back') {
return { status: 'back', projectDir: args.projectDir };
}
selectedBackend = backend;
}
const defaults = DEFAULTS[selectedBackend];
const model = defaults.model;
const dimensions = defaults.dimensions;
let credentialRef: string | undefined;
let credentialValue: string | undefined;
if (selectedBackend === 'openai') {
const credential = await chooseCredentialRef(selectedBackend, args, io, deps);
if (credential.status === 'back' && !args.embeddingBackend && args.inputMode !== 'disabled') {
selectedBackend = undefined;
continue;
}
if (credential.status !== 'ready') {
return { status: credential.status, projectDir: args.projectDir };
}
credentialRef = credential.ref;
credentialValue = credential.value;
}
let managedLocalEmbeddings: ManagedLocalEmbeddingsDaemon | undefined;
if (selectedBackend === LOCAL_EMBEDDING_BACKEND) {
const ensureLocalEmbeddings = deps.ensureLocalEmbeddings ?? ensureManagedLocalEmbeddingsDaemon;
try {
managedLocalEmbeddings = await ensureLocalEmbeddings({
cliVersion: args.cliVersion,
projectDir: args.projectDir,
installPolicy: args.runtimeInstallPolicy,
io,
});
} catch (error) {
io.stderr.write(`${error instanceof Error ? error.message : String(error)}\n`);
return { status: 'failed', projectDir: args.projectDir };
}
}
const healthConfig =
selectedBackend === LOCAL_EMBEDDING_BACKEND && managedLocalEmbeddings
? managedLocalEmbeddingHealthConfig({
baseUrl: managedLocalEmbeddings.baseUrl,
model,
dimensions,
})
: buildHealthConfig({
backend: selectedBackend,
model,
dimensions,
credentialValue,
});
const healthSpinner = (deps.spinner ?? (() => createStaticCliSpinner(io)))();
const progress = startHealthCheckProgress(healthSpinner, healthCheckStartText(selectedBackend, model, dimensions));
let health: KtxEmbeddingHealthCheckResult;
try {
health = await healthCheck(healthConfig);
} catch (error) {
progress.fail('Embedding test failed');
throw error;
}
if (health.ok) {
progress.succeed(`Embedding test passed (${model}, ${dimensions} dimensions)`);
await persistEmbeddingConfig(
args.projectDir,
selectedBackend === LOCAL_EMBEDDING_BACKEND
? {
backend: 'sentence-transformers' as const,
model,
dimensions,
}
: buildProjectEmbeddingConfig({
backend: selectedBackend,
model,
dimensions,
credentialRef,
}),
);
io.stdout.write(`│ Embeddings ready: yes (${model}, ${dimensions} dimensions)\n`);
return { status: 'ready', projectDir: args.projectDir };
}
progress.fail('Embedding test failed');
const stderrTail =
selectedBackend === 'sentence-transformers'
? await readLocalEmbeddingDaemonStderrTail(managedLocalEmbeddings?.stderrLog)
: [];
io.stderr.write(
selectedBackend === 'sentence-transformers'
? `${localEmbeddingSetupMessage(health.message, stderrTail)}\n`
: `Embedding health check failed: ${health.message}\n`,
);
if (args.inputMode === 'disabled') {
return { status: 'failed', projectDir: args.projectDir };
}
if (selectedBackend !== 'sentence-transformers' && (args.embeddingApiKeyEnv || args.embeddingApiKeyFile)) {
return { status: 'failed', projectDir: args.projectDir };
}
const nextAction =
selectedBackend === 'sentence-transformers' ? await promptAfterLocalEmbeddingFailure(deps) : 'retry';
if (nextAction === 'back') {
return { status: 'back', projectDir: args.projectDir };
}
if (nextAction === 'openai') {
selectedBackend = nextAction;
}
}
}