dataiku-sdk 4.0.0 → 4.1.0

package/skills/dataiku-dss/references/discovery.md

@@ -8,7 +8,7 @@
 - Exits: 0 success, 1 usage/configuration, 2 DSS/internal/permission-or-environment, 3 transient/retryable DSS error, 4 failed long-running result/assertion.
 - Recipe payload stdout: JSON string. With `--output PATH`: exact bytes to file, JSON string equal to `PATH` on stdout.
 - General `--fields a,b,c`: object projection, element-wise for object arrays. Dotted paths traverse nested objects; missing fields become `null`; strings/scalars pass through.
-- Compact by default: `*.list` (ids, kind, next-step fields such as recipe `inputs`/`outputs`), `job get` (no log tail), `dataset get`/`info`, `project get`; `--full`/`--fields` use DSS objects. Lists: `--contains TEXT` filters ids; `--limit N` caps (`list_truncated` warning).
+- Compact by default: `*.list` (ids, kind, next-step fields such as recipe `inputs`/`outputs`), `job get` (no log tail), `dataset get`/`info`, `project get`, `scenario get-run`/`last-runs`; `--full`/`--fields` use DSS objects. Lists: `--contains TEXT` filters ids; `--limit N` caps (`list_truncated` warning).
 
 ## Discovery
 
@@ -25,4 +25,4 @@ dss agent contract --fields commands.actions
 
 `commands run`: resource→action summary (~1.6k tokens). Unknown name: `--contains TEXT` gives the top 5 actions with usage. Known resource: `agent contract --fields commands.actions.RESOURCE`. Bootstrap once with the six-field projection (~250 tokens). Global flags (`--fields`, `--verbose`, connection/TLS) appear once in `commands.globalFlags`, not per entry.
 
-Look up syntax before invoking: `RESOURCE.ACTION` is one entry, `.FIELD` nests, commas batch, keys echo selectors. Reads: `usage,description,flags,examples`. Writes: the full entry once (flags, side effects, idempotency, dry-run, payload schemas, cleanup, exits). Unknown flags, resources, and actions exit 1 with the usage line or valid options. Full registry: `commands run --output PATH` only; stdout is `{"path":"PATH"}`.
+Look up syntax before invoking: `RESOURCE.ACTION` is one entry, `.FIELD` nests, commas batch, keys echo selectors. Unknown selectors among known ones are skipped (`unknown_fields` warning with `suggestions`). Reads: `usage,description,flags,examples`. Writes: the full entry once (flags, side effects, idempotency, dry-run, payload schemas, cleanup, exits). Unknown flags, resources, and actions exit 1 with the usage line or valid options. Full registry: `commands run --output PATH` only; stdout is `{"path":"PATH"}`.

package/skills/dataiku-dss/references/troubleshooting.md

@@ -4,6 +4,7 @@
 
 - Code/SQL: use `--file`/`--sql-file`; shells, especially PowerShell, mangle quotes, `$`, and multiline text. Non-UTF-8 consoles (e.g. Windows cp1252): write non-ASCII results to UTF-8 files or `--output PATH`, not the console.
 - Build errors: `dss job log <id> --errors-only` selects errors/tracebacks. Plain `job log` prints the last 100 lines (`log_truncated`); for full logs, use `--output PATH` without `--errors-only`. Logs may arrive as one JVM-noisy line; `Error in Python process: At line <N>` identifies the payload source line.
+- Failed scenario: `dss scenario get-run ID RUNID` gives `failure` {stepId, jobId, dataset, message} and `next` argv (`job log JOBID --errors-only`); `run-and-wait` failures carry both.
 - Schemas: code recipes set output schemas at run time. Visual/SQL query recipes: `recipe create` fills created or empty outputs (`outputSchemaUpdated`); `recipe set-payload` updates empty or DSS-derived ones, warns `recipe_output_schema_outdated` for hand-edited ones. Apply with `dss recipe update-schema NAME` (one recipe), `dss flow propagate-schema DATASET --wait` (every downstream recipe), or build `--auto-update-schema`. `built_dataset_has_no_columns`: the DONE build wrote nothing. `nothing_to_build` (exit 4): no recipe writes the managed target. Manual columns: `dss dataset refresh-schema NAME --data-file columns.json`. `dss dataset validate-build` catches file-backed misconfiguration.
 - `Inline` (editable) datasets: the public API cannot write rows, so `dataset create` rejects them; use `UploadedFiles` + `dataset upload-file`, or a python recipe output. `code run` is a scenario step; some DSS versions refuse dataset writes there.
 - `dss dataset download`: default 100k rows; result `{ path, rows, truncated, limit }`. If truncated, raise `--limit`. Specify `--output` or read the `dataset_download_default_location` path warning. CSV is spreadsheet-safe; `--raw-data` retains formula prefixes. For large tables, aggregate in SQL or use a recipe.