diff --git a/packages/context/prompts/memory_agent_bundle_ingest_work_unit.md b/packages/context/prompts/memory_agent_bundle_ingest_work_unit.md index 9645ccc4..45c0d6cd 100644 --- a/packages/context/prompts/memory_agent_bundle_ingest_work_unit.md +++ b/packages/context/prompts/memory_agent_bundle_ingest_work_unit.md @@ -1,5 +1,5 @@ -You are processing ONE WorkUnit of a multi-file ingest bundle. The WorkUnit gives you a slice of raw source files (LookML views, dbt/MetricFlow YAMLs, Metabase card JSONs, or similar) and you must translate that slice into KTX semantic-layer sources and/or knowledge wiki pages, in one pass. Prior WorkUnits in this same job may have already written SL sources and wiki pages; their writes are visible on the working branch and searchable via `wiki_sl_search`. +You are processing ONE WorkUnit of a multi-file ingest bundle. The WorkUnit gives you a slice of raw source files (LookML views, dbt/MetricFlow YAMLs, Metabase card JSONs, Notion pages, or similar) and you must translate that slice into KTX semantic-layer sources and/or knowledge wiki pages, in one pass. Prior WorkUnits in this same job may have already written SL sources and wiki pages; their writes are visible on the working branch and discoverable with `discover_data`. @@ -8,9 +8,9 @@ Assertive. The bundle was explicitly submitted for ingest. Default to capturing 1. Read this WorkUnit's section at the end of the user prompt. It lists your `rawFiles`, any unchanged `dependencyPaths` you may need to resolve references, the `peerFileIndex` (paths only; you CANNOT read them), the source's `skillNames`, and any `priorProvenance` rows telling you what earlier syncs produced from these files. -2. Load the per-source review skill first (e.g. `lookml_ingest`, `metricflow_ingest`, `dbt_ingest`), then `sl_capture` and `knowledge_capture`, and `ingest_triage` last. The triage skill tells you how to react when `wiki_sl_search` reveals that a prior WU already wrote something overlapping. +2. Load the per-source review skill first (e.g. `lookml_ingest`, `metricflow_ingest`, `dbt_ingest`), then `sl_capture` and `knowledge_capture`, and `ingest_triage` last. The triage skill tells you how to react when `discover_data` reveals that a prior WU already wrote something overlapping. 3. If the system prompt includes ``, read those pins before choosing artifact keys. A pin's `canonicalArtifactKey` is the preferred artifact for its `contestedKey`: prefer editing the pinned canonical artifact when it already exists or when this raw file clearly updates it. Do not create a duplicate contested artifact when a pin says another artifact is canonical; use a specific disambiguated key only when the raw file describes a genuinely different domain. -4. For each raw file: call `read_raw_file` (or `read_raw_span` for slicing large files) to load content. Before writing a new SL source or wiki page, call `wiki_sl_search` for each candidate name to find prior-WU writes; apply `ingest_triage` when you hit one, and apply any matching canonical pin before deciding whether to edit, rename, or skip. +4. For each raw file: call `read_raw_file` (or `read_raw_span` for slicing large files) to load content. Before writing a new SL source or wiki page, call `discover_data` for each candidate source, table, metric, or topic name to find prior-WU writes, existing wiki pages, SL sources, and raw warehouse matches; apply `ingest_triage` when you hit one, and apply any matching canonical pin before deciding whether to edit, rename, or skip. 5. For every `wiki_write`, `wiki_remove`, `sl_write_source`, or `sl_edit_source` call, include `rawPaths` with only the raw file paths that directly support that action. If one artifact synthesizes several files, list each contributing raw file. Do not include unrelated files from the same WorkUnit. 6. When `priorProvenance` names an existing artifact for one of your raw files, prefer `sl_edit` over `sl_write` for that artifact: the re-ingest change rule says expression-only changes replace silently, grain/column/filter changes replace and flag. 7. When a raw file cannot map to normal SL and you use a fallback path, call `emit_unmapped_fallback` exactly once for that raw file and reason. Use `fallback: "sql_standalone"` for a standalone SQL source, `fallback: "wiki_only"` for documentation-only capture, and `fallback: "flagged"` when no reliable artifact can be written. @@ -24,7 +24,7 @@ All wiki writes go to the GLOBAL scope. Bundle ingests are not personal. The `wi - Do not read peer files; only files listed in `rawFiles` or `dependencyPaths` are accessible. `read_raw_file` will reject everything else. - Do not invent measures/joins/rules not declared in the raw files. -- Do not invent physical column names or grain keys. For table-backed SL sources, every `columns:`, `grain:`, `joins:`, `segments:`, and `measures[].expr` column must come from raw-file column declarations or warehouse-backed discovery (`wiki_sl_search`, `sl_discover`, `sl_describe_table`). If column names are not confirmed, capture the business context in wiki instead of writing a full SL source. +- Do not invent physical column names or grain keys. For table-backed SL sources, every `columns:`, `grain:`, `joins:`, `segments:`, and `measures[].expr` column must come from raw-file column declarations or warehouse-backed discovery (`discover_data`, `sl_discover`, `entity_details`). If column names are not confirmed, capture the business context in wiki instead of writing a full SL source. - Do not write context-source overlays into the context source connection just because that is the current WorkUnit connection. Use `sl_discover` across data sources and write the SL artifact to the warehouse/data-source connection that owns the matching manifest. If there is no confirmed target connection, use `emit_unmapped_fallback` and wiki capture. - Do not duplicate an artifact that prior provenance says you already produced; update it. - Do not silently accept a name collision with a prior WU's write when the formula differs. Trigger `ingest_triage`. diff --git a/packages/context/src/ingest/adapters/notion/chunk.ts b/packages/context/src/ingest/adapters/notion/chunk.ts index 7d85fb76..bdae8622 100644 --- a/packages/context/src/ingest/adapters/notion/chunk.ts +++ b/packages/context/src/ingest/adapters/notion/chunk.ts @@ -8,7 +8,7 @@ const MAX_NOTION_WORK_UNIT_CHARS = 40_000; export const NOTION_ORG_KNOWLEDGE_WARNING = 'Anything accessible to this Notion integration can become organization knowledge.'; const NOTION_SL_WRITE_GUIDANCE = - 'Write wiki entries with wiki_write. Wiki keys must be flat slugs like orbit-company-overview, not orbit/company-overview. Search existing wiki pages for the same tables or sl_refs before creating a new page. Only write or edit SL sources after sl_discover/sl_read_source confirms a mapped non-Notion target source; if no mapped target exists, emit_unmapped_fallback and keep the fact wiki-only. Notion dataSourceCount counts Notion databases/data sources only, not warehouse/dbt mappings. If a warehouse/dbt connection exists but the named table or source is absent, use reason no_physical_table rather than no_connection_mapping. Do not create SL sources under the Notion connection just because a page mentions a warehouse table.'; + 'Write wiki entries with wiki_write. Wiki keys must be flat slugs like orbit-company-overview, not orbit/company-overview. Search existing wiki pages, SL sources, and raw warehouse schema for the same tables or sl_refs with discover_data before creating a new page. Only write or edit SL sources after discover_data plus sl_discover/sl_read_source or entity_details confirms a mapped non-Notion target source; if no mapped target exists, emit_unmapped_fallback and keep the fact wiki-only. Notion dataSourceCount counts Notion databases/data sources only, not warehouse/dbt mappings. If a warehouse/dbt connection exists but the named table or source is absent, use reason no_physical_table rather than no_connection_mapping. Do not create SL sources under the Notion connection just because a page mentions a warehouse table.'; async function walk(root: string): Promise { const entries = await readdir(root, { withFileTypes: true, recursive: true }); @@ -117,7 +117,7 @@ export async function chunkNotionStagedDir(stagedDir: string, diffSet?: DiffSet) reconcileNotes: [ `Notion maxKnowledgeCreatesPerRun=${manifest.maxKnowledgeCreatesPerRun}`, `Notion maxKnowledgeUpdatesPerRun=${manifest.maxKnowledgeUpdatesPerRun}`, - 'Notion dataSourceCount is Notion-only; use sl_discover for warehouse/dbt mapping decisions.', + 'Notion dataSourceCount is Notion-only; use discover_data/entity_details for warehouse/dbt mapping decisions.', 'Reconcile Notion wiki pages sharing tables/sl_refs before creating distinct artifacts.', ], contextReport: { diff --git a/packages/context/src/ingest/adapters/notion/notion.adapter.test.ts b/packages/context/src/ingest/adapters/notion/notion.adapter.test.ts index 76e28658..0f500d5e 100644 --- a/packages/context/src/ingest/adapters/notion/notion.adapter.test.ts +++ b/packages/context/src/ingest/adapters/notion/notion.adapter.test.ts @@ -250,6 +250,8 @@ describe('NotionSourceAdapter', () => { }); expect(result.workUnits[0].notes).toContain('Synthesize durable wiki and SL knowledge'); expect(result.workUnits[0].notes).toContain('emit_unmapped_fallback'); + expect(result.workUnits[0].notes).toContain('discover_data'); + expect(result.workUnits[0].notes).toContain('entity_details'); expect(result.workUnits[0].notes).toContain('use reason no_physical_table rather than no_connection_mapping'); expect(result.workUnits[0].notes).toContain('Do not create SL sources under the Notion connection'); expect(result.workUnits[0].notes).toContain( @@ -258,7 +260,7 @@ describe('NotionSourceAdapter', () => { expect(result.reconcileNotes).toEqual([ 'Notion maxKnowledgeCreatesPerRun=25', 'Notion maxKnowledgeUpdatesPerRun=20', - 'Notion dataSourceCount is Notion-only; use sl_discover for warehouse/dbt mapping decisions.', + 'Notion dataSourceCount is Notion-only; use discover_data/entity_details for warehouse/dbt mapping decisions.', 'Reconcile Notion wiki pages sharing tables/sl_refs before creating distinct artifacts.', ]); expect(result.contextReport).toEqual({ capped: false, warnings: [NOTION_ORG_KNOWLEDGE_WARNING] }); diff --git a/packages/context/src/ingest/ingest-prompts.test.ts b/packages/context/src/ingest/ingest-prompts.test.ts index 43985ee9..8adcbfef 100644 --- a/packages/context/src/ingest/ingest-prompts.test.ts +++ b/packages/context/src/ingest/ingest-prompts.test.ts @@ -29,6 +29,18 @@ describe('ingest prompt assets', () => { expect(prompt).not.toMatch(forbiddenProductPattern()); }); + it('uses shipped warehouse verification tools in the WorkUnit prompt', async () => { + const prompt = await readFile( + new URL('../../prompts/memory_agent_bundle_ingest_work_unit.md', import.meta.url), + 'utf-8', + ); + + expect(prompt).toContain('discover_data'); + expect(prompt).toContain('entity_details'); + expect(prompt).not.toContain('wiki_sl_search'); + expect(prompt).not.toContain('sl_describe_table'); + }); + it('does not route historic-SQL through page-triage prompt examples', async () => { const prompt = await readFile(new URL('../../prompts/skills/page_triage_classifier.md', import.meta.url), 'utf-8');