feat(cli)!: remove fast mode; ktx ingest always builds enriched context (KLO-721)

Fast mode (the ktx ingest --fast/--deep database-ingest depth toggle) is removed.
ktx ingest now always builds the full enriched ("deep") context. There is no
structural fallback: a database connection without a configured model and
embeddings fails the enrichment-readiness preflight before any work runs, with
a 'Run ktx setup to configure a model and embeddings' hint.

- Remove --fast/--deep flags, the per-connection context.depth field, and the
  ktx setup depth prompt (delete setup-database-context-depth.ts).
- Rename ingest-depth.ts -> connection-drivers.ts; ingest always requests scan
  mode 'enriched'; readiness gate (enrichmentReadinessGaps) runs for every
  database target.
- Drop the database-context-depth telemetry step (Node + Python schema mirrors
  regenerated).
- Update CLI, setup, context-build view, docs, the public ktx skill, and the
  release-smoke / artifacts scripts (now assert the no-LLM guard failure).

ktx status --fast (a separate network-probe flag) is unchanged.

Follow-ups: KLO-726 (live progress for ktx ingest --all), KLO-727 (restore
credentialed successful-ingest release smoke coverage).
This commit is contained in:
Andrey Avtomonov 2026-05-29 17:27:32 +02:00
parent 8ebc4ce107
commit e7538fb807
34 changed files with 222 additions and 884 deletions

View file

@ -228,11 +228,11 @@ describe('renderContextBuildView', () => {
const rendered = renderContextBuildView(state, {
styled: false,
warnings: ['--deep affects database ingest only; ignoring it for docs.'],
warnings: ['--query-history affects database ingest only; ignoring it for docs.'],
});
expect(rendered).toContain('Warnings:');
expect(rendered).toContain('--deep affects database ingest only; ignoring it for docs.');
expect(rendered).toContain('--query-history affects database ingest only; ignoring it for docs.');
});
it('renders public notices in the foreground view before warnings', () => {
@ -243,7 +243,6 @@ describe('renderContextBuildView', () => {
operation: 'database-ingest',
debugCommand: 'ktx ingest warehouse --debug',
steps: ['database-schema', 'query-history'],
databaseDepth: 'deep',
detectRelationships: true,
queryHistory: { enabled: true, dialect: 'postgres' },
},
@ -252,12 +251,12 @@ describe('renderContextBuildView', () => {
const rendered = renderContextBuildView(state, {
styled: false,
notices: ['Schema ingest runs before query history for warehouse.'],
warnings: ['--query-history requires deep ingest; running warehouse with --deep.'],
warnings: ['--query-history is not supported for sqlite; running schema ingest for local.'],
});
expect(rendered.indexOf('Notices:')).toBeLessThan(rendered.indexOf('Warnings:'));
expect(rendered).toContain('Schema ingest runs before query history for warehouse.');
expect(rendered).toContain('--query-history requires deep ingest; running warehouse with --deep.');
expect(rendered).toContain('--query-history is not supported for sqlite; running schema ingest for local.');
});
it('renders dynamic separator matching header width', () => {
@ -653,7 +652,6 @@ describe('runContextBuild', () => {
inputMode: 'disabled',
targetConnectionId: 'warehouse',
all: false,
depth: 'fast',
queryHistory: 'default',
},
io.io,
@ -665,7 +663,6 @@ describe('runContextBuild', () => {
expect(executeTarget.mock.calls[0]?.[0]).toMatchObject({
connectionId: 'warehouse',
operation: 'database-ingest',
databaseDepth: 'fast',
});
expect(io.stdout()).toContain('Databases:');
expect(io.stdout()).toContain('warehouse');
@ -716,7 +713,7 @@ describe('runContextBuild', () => {
it('renders localhost SQL analysis refusal as a runtime failure during query history', async () => {
const io = makeIo();
const project = projectWithConnections({
warehouse: { driver: 'postgres', context: { depth: 'deep', queryHistory: { enabled: true } } },
warehouse: { driver: 'postgres', context: { queryHistory: { enabled: true } } },
});
const executeTarget = vi.fn(async (target, _args, targetIo) => {
targetIo.stderr.write('connect ECONNREFUSED 127.0.0.1:8765\n');
@ -751,7 +748,7 @@ describe('runContextBuild', () => {
it('uses captured query-history stderr instead of generic failed-at detail', async () => {
const io = makeIo();
const project = projectWithConnections({
warehouse: { driver: 'postgres', context: { depth: 'deep', queryHistory: { enabled: true } } },
warehouse: { driver: 'postgres', context: { queryHistory: { enabled: true } } },
});
const executeTarget = vi.fn(async (target, _args, targetIo) => {
targetIo.stdout.write('KTX scan completed\n');
@ -768,7 +765,7 @@ describe('runContextBuild', () => {
operation: 'query-history',
status: 'failed',
detail:
'warehouse failed at query-history. Retry: ktx ingest warehouse --project-dir /tmp/project --deep --query-history',
'warehouse failed at query-history. Retry: ktx ingest warehouse --project-dir /tmp/project --query-history',
},
{ operation: 'source-ingest', status: 'skipped' },
{ operation: 'memory-update', status: 'skipped' },
@ -785,7 +782,7 @@ describe('runContextBuild', () => {
expect(result).toEqual({ exitCode: 1 });
expect(io.stdout()).toContain('Missing bundled Python runtime manifest: /tmp/assets/python/manifest.json.');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --deep --query-history');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --query-history');
expect(io.stdout()).not.toContain('Then retry the runtime-backed KTX command');
expect(io.stdout()).not.toContain('warehouse failed at query-history');
expect(io.stdout().match(/Retry: /g)).toHaveLength(1);
@ -899,12 +896,12 @@ describe('runContextBuild', () => {
const io = makeIo();
const project: KtxPublicIngestProject = {
...projectWithConnections({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
}),
config: {
...projectWithConnections({ warehouse: { driver: 'postgres' } }).config,
connections: {
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
},
llm: {
provider: { backend: 'gateway', gateway: { api_key: 'env:KTX_GATEWAY_API_KEY' } }, // pragma: allowlist secret

View file

@ -702,7 +702,7 @@ describe('runKtxCli', () => {
const publicIngest = vi.fn().mockResolvedValue(0);
await expect(
runKtxCli(['--project-dir', tempDir, 'ingest', 'warehouse', '--fast', '--no-input'], testIo.io, {
runKtxCli(['--project-dir', tempDir, 'ingest', 'warehouse', '--no-input'], testIo.io, {
publicIngest,
}),
).resolves.toBe(0);
@ -715,7 +715,6 @@ describe('runKtxCli', () => {
all: false,
json: false,
inputMode: 'disabled',
depth: 'fast',
queryHistory: 'default',
cliVersion,
runtimeInstallPolicy: 'never',
@ -725,12 +724,12 @@ describe('runKtxCli', () => {
expect(testIo.stderr()).toBe(`Project: ${tempDir}\n`);
});
it('routes public ingest --all --deep with JSON output', async () => {
it('routes public ingest --all with JSON output', async () => {
const testIo = makeIo();
const publicIngest = vi.fn().mockResolvedValue(0);
await expect(
runKtxCli(['--project-dir', tempDir, 'ingest', '--all', '--deep', '--json'], testIo.io, {
runKtxCli(['--project-dir', tempDir, 'ingest', '--all', '--json'], testIo.io, {
publicIngest,
}),
).resolves.toBe(0);
@ -742,7 +741,6 @@ describe('runKtxCli', () => {
all: true,
json: true,
inputMode: 'auto',
depth: 'deep',
queryHistory: 'default',
cliVersion,
runtimeInstallPolicy: 'prompt',
@ -786,20 +784,6 @@ describe('runKtxCli', () => {
expect(testIo.stderr()).toContain('Choose only one runtime install mode: --yes or --no-input');
});
it('rejects mutually exclusive public ingest depth flags before dispatch', async () => {
const testIo = makeIo();
const publicIngest = vi.fn().mockResolvedValue(0);
await expect(
runKtxCli(['--project-dir', '/tmp/project', 'ingest', 'warehouse', '--fast', '--deep'], testIo.io, {
publicIngest,
}),
).resolves.toBe(1);
expect(publicIngest).not.toHaveBeenCalled();
expect(testIo.stderr()).toMatch(/option '--(deep|fast)' cannot be used with option '--(fast|deep)'/);
});
it.each(['run', 'status', 'watch', 'replay'])(
'routes former ingest subcommand name "%s" as a connection id',
async (connectionId) => {
@ -890,8 +874,6 @@ describe('runKtxCli', () => {
expect(testIo.stdout()).toContain('Usage: ktx ingest');
expect(testIo.stdout()).toContain('Build or inspect KTX context');
expect(testIo.stdout()).toContain('--all');
expect(testIo.stdout()).toContain('--fast');
expect(testIo.stdout()).toContain('--deep');
expect(testIo.stdout()).toContain('--query-history');
expect(testIo.stdout()).toContain('--no-query-history');
expect(testIo.stdout()).toContain('--query-history-window-days <days>');

View file

@ -88,7 +88,7 @@ function deepReadyProject(
describe('buildPublicIngestPlan', () => {
it('plans warehouse connections as scan targets and source connections as source ingest targets', () => {
const project = projectWithConnections({
const project = deepReadyProject({
warehouse: { driver: 'postgres' },
prod_metabase: { driver: 'metabase', api_url: 'https://metabase.example.com' },
docs: { driver: 'notion' },
@ -103,8 +103,7 @@ describe('buildPublicIngestPlan', () => {
operation: 'database-ingest',
debugCommand: 'ktx ingest warehouse --debug',
steps: ['database-schema'],
databaseDepth: 'fast',
detectRelationships: false,
detectRelationships: true,
queryHistory: { enabled: false },
},
{
@ -139,61 +138,6 @@ describe('buildPublicIngestPlan', () => {
expect(plan.targets.map((target) => target.connectionId).sort()).toEqual(['docs', 'warehouse']);
});
it('resolves database depth from flags, stored context, and defaults', () => {
const project = projectWithConnections({
fast_default: { driver: 'postgres' },
deep_default: { driver: 'postgres', context: { depth: 'deep' } },
docs: { driver: 'notion' },
});
expect(
buildPublicIngestPlan(project, {
projectDir: '/tmp/project',
targetConnectionId: 'fast_default',
all: false,
queryHistory: 'default',
}).targets[0],
).toMatchObject({ connectionId: 'fast_default', databaseDepth: 'fast', queryHistory: { enabled: false } });
expect(
buildPublicIngestPlan(project, {
projectDir: '/tmp/project',
targetConnectionId: 'deep_default',
all: false,
queryHistory: 'default',
}).targets[0],
).toMatchObject({ connectionId: 'deep_default', databaseDepth: 'deep' });
expect(
buildPublicIngestPlan(project, {
projectDir: '/tmp/project',
targetConnectionId: 'docs',
all: false,
depth: 'deep',
queryHistory: 'default',
}).warnings,
).toEqual(['--deep affects database ingest only; ignoring it for docs.']);
});
it('does not infer deep ingest from legacy scanMode values', () => {
const project = projectWithConnections({
warehouse: { driver: 'postgres' },
});
const plan = buildPublicIngestPlan(project, {
projectDir: '/tmp/project',
targetConnectionId: 'warehouse',
all: false,
scanMode: 'enriched',
});
expect(plan.targets[0]).toMatchObject({
connectionId: 'warehouse',
databaseDepth: 'fast',
steps: ['database-schema'],
});
});
it('rejects stale local Looker source driver aliases', () => {
const project = projectWithConnections({
local_looker: { driver: 'local_looker' } as never,
@ -204,8 +148,8 @@ describe('buildPublicIngestPlan', () => {
);
});
it('upgrades effective depth when query history is explicitly enabled', () => {
const project = projectWithConnections({
it('enables query history when explicitly requested even if stored config disables it', () => {
const project = deepReadyProject({
warehouse: { driver: 'postgres', context: { queryHistory: { enabled: false } } },
});
@ -213,17 +157,16 @@ describe('buildPublicIngestPlan', () => {
projectDir: '/tmp/project',
targetConnectionId: 'warehouse',
all: false,
depth: 'fast',
queryHistory: 'enabled',
queryHistoryWindowDays: 30,
});
expect(plan.targets[0]).toMatchObject({
connectionId: 'warehouse',
databaseDepth: 'deep',
queryHistory: { enabled: true, windowDays: 30, dialect: 'postgres' },
steps: ['database-schema', 'query-history'],
});
expect(plan.warnings).toEqual(['--query-history requires deep ingest; running warehouse with --deep.']);
expect(plan.warnings).toEqual([]);
});
it('warns and skips query history for unsupported database drivers', () => {
@ -238,7 +181,6 @@ describe('buildPublicIngestPlan', () => {
expect(plan.targets[0]).toMatchObject({
connectionId: 'local',
databaseDepth: 'fast',
queryHistory: { enabled: false, unsupported: true },
});
expect(plan.warnings).toEqual(['--query-history is not supported for sqlite; running schema ingest for local.']);
@ -249,12 +191,11 @@ describe('buildPublicIngestPlan', () => {
deepReadyProject({
local: { driver: 'sqlite' },
mysql_warehouse: { driver: 'mysql' },
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
}),
{
projectDir: '/tmp/project',
all: true,
depth: 'deep',
queryHistory: 'enabled',
},
);
@ -326,7 +267,6 @@ describe('buildPublicIngestPlan', () => {
expect(plan.targets[0]).toMatchObject({
connectionId: 'warehouse',
databaseDepth: 'deep',
queryHistory: { enabled: true, dialect: 'postgres', windowDays: 30 },
steps: ['database-schema', 'query-history'],
});
@ -334,7 +274,7 @@ describe('buildPublicIngestPlan', () => {
it('adds a schema-first notice when query history is explicitly enabled', () => {
const project = deepReadyProject({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
expect(
@ -363,34 +303,15 @@ describe('buildPublicIngestPlan', () => {
expect(plan.targets[0]).toMatchObject({
connectionId: 'local',
databaseDepth: 'fast',
queryHistory: { enabled: false, windowDays: 30, unsupported: true },
steps: ['database-schema'],
});
expect(plan.warnings).toEqual(['--query-history is not supported for sqlite; running schema ingest for local.']);
});
it('aggregates ignored database-depth warnings for all source targets', () => {
const plan = buildPublicIngestPlan(
projectWithConnections({
warehouse: { driver: 'postgres' },
docs: { driver: 'notion' },
dbt: { driver: 'dbt' },
}),
{
projectDir: '/tmp/project',
all: true,
depth: 'deep',
queryHistory: 'default',
},
);
expect(plan.warnings).toEqual(['--deep ignored for 2 non-database sources.']);
});
it('records a preflight failure for deep database ingest when readiness config is missing', () => {
it('records a preflight failure for database ingest when enrichment readiness config is missing', () => {
const project = projectWithConnections({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
const plan = buildPublicIngestPlan(project, {
@ -402,15 +323,14 @@ describe('buildPublicIngestPlan', () => {
expect(plan.targets[0]).toMatchObject({
connectionId: 'warehouse',
databaseDepth: 'deep',
preflightFailure:
'warehouse requires deep ingest readiness: model configuration, scan enrichment mode, scan embeddings. Run ktx setup or rerun with --fast.',
'warehouse cannot be ingested: enrichment is not configured (model configuration, scan enrichment mode, scan embeddings). Run ktx setup to configure a model and embeddings.',
});
});
it('honors scan.relationships.enabled when planning deep database ingest', () => {
it('honors scan.relationships.enabled when planning database ingest', () => {
const plan = buildPublicIngestPlan(
deepReadyProject({ warehouse: { driver: 'postgres', context: { depth: 'deep' } } }, false),
deepReadyProject({ warehouse: { driver: 'postgres' } }, false),
{
projectDir: '/tmp/project',
targetConnectionId: 'warehouse',
@ -421,7 +341,6 @@ describe('buildPublicIngestPlan', () => {
expect(plan.targets[0]).toMatchObject({
connectionId: 'warehouse',
databaseDepth: 'deep',
detectRelationships: false,
});
});
@ -432,11 +351,11 @@ describe('runKtxPublicIngest', () => {
vi.unstubAllEnvs();
});
it('maps fast and deep database targets to scan internals', async () => {
it('maps database targets to enriched scan internals', async () => {
const io = makeIo();
const project = deepReadyProject({
fast: { driver: 'postgres' },
deep: { driver: 'postgres', context: { depth: 'deep' } },
first: { driver: 'postgres' },
second: { driver: 'postgres' },
});
const runScan = vi.fn(async () => 0);
@ -450,12 +369,12 @@ describe('runKtxPublicIngest', () => {
expect(runScan).toHaveBeenNthCalledWith(
1,
expect.objectContaining({ connectionId: 'deep', mode: 'enriched', detectRelationships: true }),
expect.objectContaining({ connectionId: 'first', mode: 'enriched', detectRelationships: true }),
expect.anything(),
);
expect(runScan).toHaveBeenNthCalledWith(
2,
expect.objectContaining({ connectionId: 'fast', mode: 'structural', detectRelationships: false }),
expect.objectContaining({ connectionId: 'second', mode: 'enriched', detectRelationships: true }),
expect.anything(),
);
});
@ -467,7 +386,7 @@ describe('runKtxPublicIngest', () => {
try {
await initKtxProject({ projectDir });
const io = makeIo({ isTTY: true });
const project = projectWithConnections({
const project = deepReadyProject({
warehouse: { driver: 'sqlite', path: join(projectDir, 'warehouse.sqlite') },
});
@ -614,7 +533,7 @@ describe('runKtxPublicIngest', () => {
it('prints the schema-first notice for explicit query-history runs', async () => {
const io = makeIo();
const project = deepReadyProject({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
const runScan = vi.fn(async () => 0);
const runIngest = vi.fn(async () => 0);
@ -640,7 +559,7 @@ describe('runKtxPublicIngest', () => {
it('suppresses internal scan output for public database ingest summaries', async () => {
const io = makeIo();
const project = projectWithConnections({ warehouse: { driver: 'postgres' } });
const project = deepReadyProject({ warehouse: { driver: 'postgres' } });
const runScan = vi.fn(async (_args, scanIo) => {
scanIo.stdout.write('KTX scan completed\n');
scanIo.stdout.write('Mode: structural\n');
@ -674,7 +593,7 @@ describe('runKtxPublicIngest', () => {
it('sanitizes captured database scan failure details in direct public output', async () => {
const io = makeIo();
const project = deepReadyProject({ warehouse: { driver: 'postgres', context: { depth: 'deep' } } });
const project = deepReadyProject({ warehouse: { driver: 'postgres' } });
const runScan = vi.fn(async (_args, scanIo) => {
scanIo.stdout.write('KTX scan enrichment failed after structural scan completed: embedding service timed out\n');
return 1;
@ -689,7 +608,6 @@ describe('runKtxPublicIngest', () => {
all: false,
json: false,
inputMode: 'disabled',
depth: 'deep',
},
io.io,
{ loadProject: vi.fn(async () => project), runScan },
@ -699,7 +617,7 @@ describe('runKtxPublicIngest', () => {
expect(io.stdout()).toContain(
'warehouse failed: Database enrichment failed after schema context completed: embedding service timed out.',
);
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --deep');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project');
expect(io.stdout()).not.toContain('KTX scan enrichment failed');
expect(io.stdout()).not.toContain('structural scan');
});
@ -743,7 +661,7 @@ describe('runKtxPublicIngest', () => {
it('suppresses historic-sql report output during direct public query-history ingest', async () => {
const io = makeIo();
const project = deepReadyProject({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
const runScan = vi.fn(async () => 0);
const runIngest = vi.fn(async (_args, ingestIo) => {
@ -794,7 +712,6 @@ describe('runKtxPublicIngest', () => {
all: false,
json: false,
inputMode: 'auto',
depth: 'fast',
queryHistory: 'default',
},
io.io,
@ -809,7 +726,6 @@ describe('runKtxPublicIngest', () => {
targetConnectionId: 'warehouse',
all: false,
entrypoint: 'ingest',
depth: 'fast',
queryHistory: 'default',
}),
io.io,
@ -821,7 +737,7 @@ describe('runKtxPublicIngest', () => {
const io = makeIo({ isTTY: true, interactive: true });
const calls: string[] = [];
const project = projectWithConnections({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
const ensureRuntime = vi.fn(async (): Promise<ManagedPythonCommandRuntime> => {
calls.push('runtime');
@ -923,10 +839,13 @@ describe('runKtxPublicIngest', () => {
it('runs all independent targets and reports partial failures', async () => {
const io = makeIo();
const project = projectWithConnections({
warehouse: { driver: 'postgres' },
prod_metabase: { driver: 'metabase', api_url: 'https://metabase.example.com' },
});
const project = deepReadyProject(
{
warehouse: { driver: 'postgres' },
prod_metabase: { driver: 'metabase', api_url: 'https://metabase.example.com' },
},
false,
);
const runScan = vi.fn(async () => 1);
const runIngest = vi.fn(async () => 0);
@ -959,7 +878,7 @@ describe('runKtxPublicIngest', () => {
command: 'run',
projectDir: '/tmp/project',
connectionId: 'warehouse',
mode: 'structural',
mode: 'enriched',
detectRelationships: false,
dryRun: false,
},
@ -967,14 +886,14 @@ describe('runKtxPublicIngest', () => {
);
expect(io.stdout()).toContain('Ingest finished with partial failures');
expect(io.stdout()).toContain('warehouse failed at database-schema.');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --fast');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project');
expect(io.stdout()).not.toContain('Debug:');
});
it('skips the query-history facet but keeps the target green when query-history fails', async () => {
const io = makeIo();
const project = deepReadyProject({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
const runScan = vi.fn(async () => 0);
const runIngest = vi.fn(async (_args, ingestIo) => {
@ -1007,14 +926,14 @@ describe('runKtxPublicIngest', () => {
'Query history failed for 60 tasks. First failure: Google Cloud authentication failed while analyzing query history',
);
expect(io.stdout()).not.toContain('warehouse failed: Error:');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --deep --query-history');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --query-history');
expect(io.stdout()).not.toContain('historic-sql');
});
it('prints the runtime artifact build hint for missing query-history runtime assets', async () => {
const io = makeIo();
const project = deepReadyProject({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
});
const runScan = vi.fn(async () => 0);
const runIngest = vi.fn(async (_args, ingestIo) => {
@ -1045,14 +964,14 @@ describe('runKtxPublicIngest', () => {
expect(io.stdout()).toContain(
'In a source checkout, build the local runtime assets with: pnpm run artifacts:build',
);
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --deep --query-history');
expect(io.stdout()).toContain('Retry: ktx ingest warehouse --project-dir /tmp/project --query-history');
expect(io.stdout()).not.toContain('Then retry the runtime-backed KTX command');
});
it('fails deep-readiness targets before work starts while continuing independent --all targets', async () => {
it('fails enrichment-readiness targets before work starts while continuing independent --all targets', async () => {
const io = makeIo();
const project = projectWithConnections({
warehouse: { driver: 'postgres', context: { depth: 'deep' } },
warehouse: { driver: 'postgres' },
docs: { driver: 'notion' },
});
const runScan = vi.fn(async () => 0);
@ -1071,12 +990,12 @@ describe('runKtxPublicIngest', () => {
expect.objectContaining({ command: 'run', connectionId: 'docs', adapter: 'notion' }),
expect.anything(),
);
expect(io.stdout()).toContain('warehouse requires deep ingest readiness');
expect(io.stdout()).toContain('warehouse cannot be ingested: enrichment is not configured');
});
it('does not infer enriched relationship scans from legacy scanMode values', async () => {
it('drives scan relationship detection from project config, not from legacy args', async () => {
const io = makeIo();
const project = deepReadyProject({ warehouse: { driver: 'postgres' } });
const project = deepReadyProject({ warehouse: { driver: 'postgres' } }, false);
const runScan = vi.fn(async () => 0);
await expect(
@ -1103,7 +1022,7 @@ describe('runKtxPublicIngest', () => {
command: 'run',
projectDir: '/tmp/project',
connectionId: 'warehouse',
mode: 'structural',
mode: 'enriched',
detectRelationships: false,
dryRun: false,
},
@ -1113,7 +1032,7 @@ describe('runKtxPublicIngest', () => {
it('prints stable JSON results', async () => {
const io = makeIo();
const project = projectWithConnections({ warehouse: { driver: 'postgres' } });
const project = deepReadyProject({ warehouse: { driver: 'postgres' } });
await expect(
runKtxPublicIngest(

View file

@ -1,7 +1,7 @@
import { mkdir, mkdtemp, readFile, rm, writeFile } from 'node:fs/promises';
import { tmpdir } from 'node:os';
import { join } from 'node:path';
import { buildDefaultKtxProjectConfig, parseKtxProjectConfig, serializeKtxProjectConfig, type KtxProjectConfig } from '../src/context/project/config.js';
import { buildDefaultKtxProjectConfig, serializeKtxProjectConfig, type KtxProjectConfig } from '../src/context/project/config.js';
import { readKtxSetupState, writeKtxSetupState } from '../src/context/project/setup-config.js';
import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest';
@ -49,7 +49,7 @@ async function writeReadyProject(projectDir: string, overrides: ReadyProjectOver
...defaults,
setup: { database_connection_ids: ['warehouse'] },
connections: {
warehouse: { driver: 'postgres', url: 'env:DATABASE_URL', context: { depth: 'deep' } },
warehouse: { driver: 'postgres', url: 'env:DATABASE_URL' },
docs: { driver: 'notion', auth_token_ref: 'env:NOTION_TOKEN', crawl_mode: 'all_accessible' },
},
llm: {
@ -407,130 +407,10 @@ describe('setup context build state', () => {
expect(io.stdout()).not.toContain('Existing context artifacts were found from setup ingest.');
});
it('treats fast database context as ready from schema manifest shards without AI artifacts', async () => {
it('requires completed relationships for database context when relationship discovery is enabled', async () => {
await writeReadyProject(tempDir, {
connections: {
warehouse: { driver: 'postgres', readonly: true, context: { depth: 'fast' } },
},
llm: { provider: { backend: 'none' }, models: {} },
scan: { enrichment: { mode: 'none' } },
});
await mkdir(join(tempDir, 'semantic-layer', 'warehouse', '_schema'), { recursive: true });
await writeFile(join(tempDir, 'semantic-layer', 'warehouse', '_schema', 'public.yaml'), 'tables: {}\n');
await writeScanReport(tempDir, '2026-05-09T10:00:00.000Z', {
mode: 'structural',
tableDescriptions: 'skipped',
columnDescriptions: 'skipped',
embeddings: 'skipped',
manifestShards: ['semantic-layer/warehouse/_schema/public.yaml'],
});
const io = makeIo();
const runContextBuildMock = vi.fn<NonNullable<KtxSetupContextDeps['runContextBuild']>>(async () => ({
exitCode: 0,
}));
await expect(
runKtxSetupContextStep(
{ projectDir: tempDir, inputMode: 'disabled' },
io.io,
{
runContextBuild: runContextBuildMock,
},
),
).resolves.toMatchObject({ status: 'ready' });
expect(runContextBuildMock).not.toHaveBeenCalled();
expect(io.stdout()).toContain('Existing context artifacts were found from setup ingest.');
});
it('stores fast context depth non-interactively when deep readiness is missing', async () => {
await writeReadyProject(tempDir, {
connections: { warehouse: { driver: 'postgres', readonly: true } },
llm: { provider: { backend: 'none' }, models: {} },
scan: { enrichment: { mode: 'none' } },
});
const io = makeIo();
const runContextBuildMock = vi.fn<NonNullable<KtxSetupContextDeps['runContextBuild']>>(async () => ({
exitCode: 0,
}));
const verifyContextReady = vi.fn(async () => ({
ready: true,
agentContextReady: true,
semanticSearchReady: true,
details: ['ready'],
}));
await expect(
runKtxSetupContextStep(
{ projectDir: tempDir, inputMode: 'disabled' },
io.io,
{ runContextBuild: runContextBuildMock, verifyContextReady },
),
).resolves.toMatchObject({ status: 'ready' });
const config = parseKtxProjectConfig(await readFile(join(tempDir, 'ktx.yaml'), 'utf-8'));
expect(config.connections.warehouse.context).toMatchObject({ depth: 'fast' });
expect(runContextBuildMock).toHaveBeenCalledWith(
expect.anything(),
expect.objectContaining({ projectDir: tempDir, inputMode: 'disabled' }),
expect.anything(),
expect.anything(),
);
expect(runContextBuildMock.mock.calls[0]?.[1]).not.toMatchObject({
scanMode: 'enriched',
detectRelationships: true,
});
});
it('prompts for database context depth after final readiness is known', async () => {
await writeReadyProject(tempDir, {
connections: { warehouse: { driver: 'postgres', readonly: true } },
llm: {
provider: { backend: 'gateway', gateway: { api_key: 'env:KTX_GATEWAY_API_KEY' } }, // pragma: allowlist secret
models: { default: 'gpt-test' },
},
scan: {
enrichment: {
mode: 'llm',
embeddings: { backend: 'openai', model: 'text-embedding-3-small', dimensions: 1536 },
},
},
});
const io = makeIo();
const select = vi.fn(async () => 'deep');
const runContextBuildMock = vi.fn(async () => ({ exitCode: 0 }));
const verifyContextReady = vi.fn(async () => ({
ready: true,
agentContextReady: true,
semanticSearchReady: true,
details: ['ready'],
}));
await expect(
runKtxSetupContextStep(
{ projectDir: tempDir, inputMode: 'auto' },
io.io,
{
prompts: { select, cancel: vi.fn() },
runContextBuild: runContextBuildMock,
verifyContextReady,
},
),
).resolves.toMatchObject({ status: 'ready' });
expect(select).toHaveBeenCalledWith(
expect.objectContaining({
message: expect.stringContaining('How much database context should KTX build?'),
}),
);
const config = parseKtxProjectConfig(await readFile(join(tempDir, 'ktx.yaml'), 'utf-8'));
expect(config.connections.warehouse.context).toMatchObject({ depth: 'deep' });
});
it('requires completed relationships for deep context when relationship discovery is enabled', async () => {
await writeReadyProject(tempDir, {
connections: {
warehouse: { driver: 'postgres', readonly: true, context: { depth: 'deep' } },
warehouse: { driver: 'postgres', readonly: true },
},
scan: { relationships: { enabled: true } },
});
@ -560,10 +440,10 @@ describe('setup context build state', () => {
expect(runContextBuildMock).toHaveBeenCalledOnce();
});
it('does not require relationships for deep context when relationship discovery is disabled', async () => {
it('does not require relationships for database context when relationship discovery is disabled', async () => {
await writeReadyProject(tempDir, {
connections: {
warehouse: { driver: 'postgres', readonly: true, context: { depth: 'deep' } },
warehouse: { driver: 'postgres', readonly: true },
},
scan: { relationships: { enabled: false } },
});
@ -620,7 +500,7 @@ describe('setup context build state', () => {
it('starts a fresh foreground build when stale state is found', async () => {
await writeReadyProject(tempDir, {
connections: { warehouse: { driver: 'postgres', readonly: true, context: { depth: 'fast' } } },
connections: { warehouse: { driver: 'postgres', readonly: true } },
});
await writeKtxSetupContextState(tempDir, {
runId: 'setup-context-local-stale',

View file

@ -262,48 +262,6 @@ describe('setup databases step', () => {
expect(prompts.select).toHaveBeenCalledTimes(1);
});
it('preserves context.depth when editing an existing database connection', async () => {
await writeFile(
join(tempDir, 'ktx.yaml'),
[
'connections:',
' warehouse:',
' driver: sqlite',
' path: ./warehouse.sqlite',
' context:',
' depth: deep',
'',
].join('\n'),
'utf-8',
);
const prompts = makePromptAdapter({
selectValues: ['edit', 'warehouse', 'continue'],
textValues: ['./warehouse.sqlite'],
});
const testConnection = vi.fn(async () => 0);
const scanConnection = vi.fn(async () => 0);
const io = makeIo();
const result = await runKtxSetupDatabasesStep(
{
projectDir: tempDir,
inputMode: 'auto',
skipDatabases: false,
databaseSchemas: [],
disableQueryHistory: true,
},
io.io,
{ prompts, testConnection, scanConnection },
);
expect(result.status, io.stderr()).toBe('ready');
const config = parseKtxProjectConfig(await readFile(join(tempDir, 'ktx.yaml'), 'utf-8'));
expect(config.connections.warehouse).toMatchObject({
driver: 'sqlite',
path: './warehouse.sqlite',
context: { depth: 'deep' },
});
});
it('labels existing database connections with the database type', async () => {
await writeFile(
join(tempDir, 'ktx.yaml'),
@ -376,7 +334,6 @@ describe('setup databases step', () => {
expect(config.connections['postgres-warehouse']).toEqual({
driver: 'postgres',
url: 'env:DATABASE_URL',
context: { depth: 'fast' },
});
});
@ -1558,7 +1515,7 @@ describe('setup databases step', () => {
);
expect(io.stdout()).not.toContain('Tables: 2');
expect(io.stdout()).toContain('◇ Building schema context for postgres-warehouse');
expect(io.stdout()).toContain('│ Running fast database ingest…');
expect(io.stdout()).toContain('│ Running database scan…');
expect(io.stdout()).toContain('◇ Schema context complete for postgres-warehouse');
expect(io.stdout()).toContain('│ Changes: 2 new tables');
expect(io.stdout()).toContain('◇ Database ready');
@ -1907,7 +1864,7 @@ describe('setup databases step', () => {
driver: 'postgres',
url: 'env:DATABASE_URL',
schemas: ['public'],
context: { queryHistory: { enabled: false }, depth: 'fast' },
context: { queryHistory: { enabled: false } },
});
expect(config.setup).toEqual({
database_connection_ids: ['warehouse'],
@ -1946,7 +1903,6 @@ describe('setup databases step', () => {
expect(config.connections.warehouse).toEqual({
driver: 'sqlite',
path: './warehouse.sqlite',
context: { depth: 'fast' },
});
expect(config.setup).toEqual({
database_connection_ids: ['warehouse'],
@ -2023,11 +1979,11 @@ describe('setup databases step', () => {
const config = parseKtxProjectConfig(await readFile(join(tempDir, 'ktx.yaml'), 'utf-8'));
expect(config.connections.warehouse).toMatchObject({ driver: 'postgres', url: 'env:DATABASE_URL' });
expect(await readFile(join(tempDir, 'ktx.yaml'), 'utf-8')).not.toContain('completed_steps:');
expect(io.stderr()).toContain('Fast database ingest failed for warehouse.');
expect(io.stderr()).toContain('│ Fast database ingest failed for warehouse.');
expect(io.stderr()).toContain(`Debug command: ktx ingest warehouse --project-dir ${tempDir} --fast --debug`);
expect(io.stderr()).toContain('Database scan failed for warehouse.');
expect(io.stderr()).toContain('│ Database scan failed for warehouse.');
expect(io.stderr()).toContain(`Debug command: ktx ingest warehouse --project-dir ${tempDir} --debug`);
expect(io.stderr()).not.toContain('Structural scan failed for warehouse.');
expect(io.stderr()).not.toMatch(/^Fast database ingest failed for warehouse\./m);
expect(io.stderr()).not.toMatch(/^Database scan failed for warehouse\./m);
});
it('prints the native SQLite rebuild command when scanning hits a Node ABI mismatch', async () => {
@ -2066,7 +2022,7 @@ describe('setup databases step', () => {
expect(io.stderr()).toContain('Native SQLite is built for a different Node.js ABI.');
expect(io.stderr()).toContain('│ Native SQLite is built for a different Node.js ABI.');
expect(io.stderr()).toContain('Fix: pnpm run native:rebuild');
expect(io.stderr()).toContain(`Retry: ktx ingest warehouse --project-dir ${tempDir} --fast`);
expect(io.stderr()).toContain(`Retry: ktx ingest warehouse --project-dir ${tempDir}`);
expect(io.stderr()).not.toContain('ktx scan');
expect(io.stderr()).not.toContain('npm rebuild');
expect(io.stderr()).not.toMatch(/^Native SQLite is built for a different Node.js ABI\./m);
@ -2364,7 +2320,7 @@ describe('setup databases step', () => {
'utf-8',
);
const io = makeIo();
const prompts = makePromptAdapter({ selectValues: ['yes', 'deep'] });
const prompts = makePromptAdapter({ selectValues: ['yes'] });
const runner = fakeHistoricSqlRunner('postgres', 'pg_stat_statements');
const historicSqlReadinessProbe = vi.fn(async () => ({
ok: true as const,
@ -2399,12 +2355,6 @@ describe('setup databases step', () => {
{ value: 'back', label: 'Back' },
],
});
expect(prompts.select).toHaveBeenNthCalledWith(
2,
expect.objectContaining({
message: expect.stringContaining('How much database context should KTX build?'),
}),
);
expect(historicSqlReadinessProbe).toHaveBeenCalledWith(
expect.objectContaining({
projectDir: tempDir,
@ -2420,7 +2370,6 @@ describe('setup databases step', () => {
minExecutions: 5,
filters: { dropTrivialProbes: true },
},
depth: 'deep',
},
});
});

View file

@ -185,7 +185,7 @@ describe('standalone built ktx CLI smoke', () => {
expect([0, 1]).toContain(result.code);
});
it('runs fast public database ingest through the built binary with manifest artifacts', async () => {
it('blocks public database ingest through the built binary when enrichment is not configured', async () => {
const projectDir = join(tempDir, 'database-ingest-project');
const init = await runSetupNewProject(projectDir);
expectSetupStderr(init);
@ -200,19 +200,10 @@ describe('standalone built ktx CLI smoke', () => {
expect(connectionTest.stdout).toContain('Driver: sqlite');
expect(connectionTest.stdout).toContain('Status: ok');
const ingest = await runBuiltCli(['ingest', 'warehouse', '--project-dir', projectDir, '--fast', '--no-input']);
expectProjectStderr(ingest, projectDir);
expect(ingest.stdout).toContain('Ingest finished');
expect(ingest.stdout).toContain('warehouse');
expect(ingest.stdout).toContain('Database schema');
expect(ingest.stdout).toContain('warehouse done');
const ingest = await runBuiltCli(['ingest', 'warehouse', '--project-dir', projectDir, '--no-input']);
expect(ingest.code).toBe(1);
expect(ingest.stdout).toContain('warehouse cannot be ingested: enrichment is not configured');
expect(ingest.stdout).not.toContain('KTX scan completed');
const manifest = await readFile(join(projectDir, 'semantic-layer/warehouse/_schema/public.yaml'), 'utf-8');
expect(manifest).toContain('customers:');
expect(manifest).toContain('orders:');
expect(manifest).toContain('source: formal');
expect(manifest).not.toContain('ai:');
}, 30_000);
it('parses gateway LLM config and OpenAI enrichment embeddings used by standalone scans without network calls', async () => {