@eigenpal/cli 0.23.6 → 0.23.8

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -570,7 +570,6 @@ Workflow and eval references:
570
570
 
571
571
  - [`reference/workflow-yaml.md`](reference/workflow-yaml.md)
572
572
  - [`reference/step-types.md`](reference/step-types.md)
573
- - [`reference/step-exec.md`](reference/step-exec.md)
574
573
  - [`reference/dataset-format.md`](reference/dataset-format.md)
575
574
  - [`reference/evaluators.md`](reference/evaluators.md)
576
575
  - [`reference/debugging.md`](reference/debugging.md)
@@ -78,9 +78,7 @@ eigenpal
78
78
  │ │ ├── list|ls
79
79
  │ │ └── get <type>
80
80
  │ ├── validate [path]
81
- │ ├── clear-local [examples...]
82
- │ └── step
83
- │ └── exec <type>
81
+ │ └── clear-local [examples...]
84
82
  ├── agents
85
83
  │ ├── list|ls
86
84
  │ ├── delete <agent-id-or-slug>
@@ -15,7 +15,6 @@ Manage workflows: push, pull, and evaluate.
15
15
  - [Versions](#versions)
16
16
  - [Step-type](#step-type)
17
17
  - [Evaluator-type](#evaluator-type)
18
- - [Step](#step)
19
18
  - [Details](#details)
20
19
  - [`eigenpal workflow list|ls [options]`](#eigenpal-workflow-listls-options)
21
20
  - [`eigenpal workflow get [options] <workflow-id>`](#eigenpal-workflow-get-options-workflow-id)
@@ -72,7 +71,6 @@ Manage workflows: push, pull, and evaluate.
72
71
  - [`eigenpal workflow step-type get [options] <type>`](#eigenpal-workflow-step-type-get-options-type)
73
72
  - [`eigenpal workflow evaluator-type list|ls [options]`](#eigenpal-workflow-evaluator-type-listls-options)
74
73
  - [`eigenpal workflow evaluator-type get [options] <type>`](#eigenpal-workflow-evaluator-type-get-options-type)
75
- - [`eigenpal workflow step exec [options] <type>`](#eigenpal-workflow-step-exec-options-type)
76
74
 
77
75
  ## Surface
78
76
 
@@ -142,9 +140,7 @@ workflow
142
140
  │ ├── list|ls
143
141
  │ └── get <type>
144
142
  ├── validate [path]
145
- ├── clear-local [examples...]
146
- └── step
147
- └── exec <type>
143
+ └── clear-local [examples...]
148
144
  ```
149
145
 
150
146
  ## Commands
@@ -249,12 +245,6 @@ workflow
249
245
  | `eigenpal workflow evaluator-type list\|ls [options]` | List every evaluator type with a one-line description. |
250
246
  | `eigenpal workflow evaluator-type get [options] <type>` | Fetch the JSON Schema for one evaluator type. Pipe through `jq` to inspect specific fields. |
251
247
 
252
- ### Step
253
-
254
- | Command | Description |
255
- | ---------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------ |
256
- | `eigenpal workflow step exec [options] <type>` | DISABLED — local mimic runners removed pending server-side redesign (EIG-104). Use `run` or `workflow experiment run` instead. |
257
-
258
248
  ## Details
259
249
 
260
250
  ### `eigenpal workflow list|ls [options]`
@@ -1276,22 +1266,3 @@ Fetch the JSON Schema for one evaluator type. Pipe through `jq` to inspect speci
1276
1266
  | Flag | Required | Default | Description |
1277
1267
  | -------------- | -------- | ------- | ------------- |
1278
1268
  | `--out <path>` | no | | Write to file |
1279
-
1280
- ### `eigenpal workflow step exec [options] <type>`
1281
-
1282
- DISABLED — local mimic runners removed pending server-side redesign (EIG-104). Use `run` or `workflow experiment run` instead.
1283
-
1284
- ### Arguments
1285
-
1286
- | Name | Required | Variadic | Description |
1287
- | ------ | -------- | -------- | ----------- |
1288
- | `type` | yes | no | |
1289
-
1290
- ### Options
1291
-
1292
- | Flag | Required | Default | Description |
1293
- | ------------------------ | -------- | ------- | -------------------------------------- |
1294
- | `--config-json <json>` | no | | (unused; kept for back-compat parsing) |
1295
- | `--config-file <path>` | no | | (unused) |
1296
- | `--inputs <pairs...>` | no | | (unused) |
1297
- | `--output-schema <path>` | no | | (unused) |
@@ -394,16 +394,16 @@ agent automations.
394
394
  <!-- GENERATED:DATASET_REFERENCE START -->
395
395
  ## Schema reference
396
396
 
397
- _Generated from `@eigenpal/types/src/eval/dataset-archive.ts`, `expected-error.ts`, and `scoped-file-ref.ts`. Do not hand-edit between the GENERATED fences — run `bun run --cwd packages/cli generate:skill`._
397
+ _Generated from `@eigenpal/types/src/eval/dataset-archive.ts`, `expected-error.ts`, and `scoped-file-ref.ts`. Do not hand-edit between the GENERATED fences: run `bun run --cwd packages/cli generate:skill`._
398
398
 
399
399
  ### Archive layout (canonical)
400
400
 
401
401
  ```text
402
- examples/<name>/input.json REQUIRED — full run input object
403
- examples/<name>/input/<file> OPTIONAL — referenced from input.json
404
- examples/<name>/expected.json OPTIONAL — success output or `{ "$error": ... }`
405
- examples/<name>/expected/<file> OPTIONAL — referenced from expected.json
406
- examples/<name>/meta.json OPTIONAL — see `DatasetMetaSchema` below
402
+ examples/<name>/input.json REQUIRED: full run input object
403
+ examples/<name>/input/<file> OPTIONAL: referenced from input.json
404
+ examples/<name>/expected.json OPTIONAL: success output or `{ "$error": ... }`
405
+ examples/<name>/expected/<file> OPTIONAL: referenced from expected.json
406
+ examples/<name>/meta.json OPTIONAL: see `DatasetMetaSchema` below
407
407
  ```
408
408
 
409
409
  Importers reject any archive containing a top-level `manifest.json` (legacy layout). Example folder names must match:
@@ -419,11 +419,11 @@ Importers reject any archive containing a top-level `manifest.json` (legacy layo
419
419
 
420
420
  Input files live under `input/` and are referenced as `{ "$file": "input/<path>" }`. Expected files use the `expected/` prefix in `expected.json`.
421
421
 
422
- ### `expected.json` — success-expected output
422
+ ### `expected.json`: success-expected output
423
423
 
424
424
  When grading success, `expected.json` is a JSON object mirroring the workflow `output:` shape. File values use `{ "$file": "expected/<path>" }`.
425
425
 
426
- ### `expected.json` — failure-expected (`$error`)
426
+ ### `expected.json`: failure-expected (`$error`)
427
427
 
428
428
  Failure-expected examples store a single top-level `$error` object. At least one of `code`, `messageContains`, or `step` is required:
429
429
 
@@ -448,8 +448,8 @@ Failure-expected examples store a single top-level `$error` object. At least one
448
448
 
449
449
  Allowed values: `"append"` \| `"replace"`
450
450
 
451
- - `append` — add examples to the existing dataset.
452
- - `replace` — wipe existing examples and import the archive fresh.
451
+ - `append`: add examples to the existing dataset.
452
+ - `replace`: wipe existing examples and import the archive fresh.
453
453
 
454
454
  ### Validation rules enforced at import
455
455
 
@@ -215,7 +215,7 @@ eigenpal runs cancel <exec-id> --json # raw server payload
215
215
 
216
216
  The worker finishes the current step, then stops — the cancellation
217
217
  check happens between every step transition, including inside
218
- `control.foreach` / `control.parallel` / `control.block` bodies.
218
+ `control.foreach` / `control.parallel` bodies.
219
219
  Behavior by current state:
220
220
 
221
221
  - `created` / `pending` — transitions straight to `cancelled`.
@@ -311,9 +311,9 @@ LLM-judge scores have variance. To reduce flake:
311
311
  - For high-stakes thresholds, run the experiment ≥3 times and average.
312
312
 
313
313
  <!-- GENERATED:EVALUATOR_CATALOG START -->
314
- ## Evaluator types — full reference
314
+ ## Evaluator types: full reference
315
315
 
316
- _Generated from `EvalConfigYamlSchema` in `@eigenpal/types/src/eval/evaluator-config.ts`. Do not hand-edit between the GENERATED fences — run `bun run --cwd packages/cli generate:skill`._
316
+ _Generated from `EvalConfigYamlSchema` in `@eigenpal/types/src/eval/evaluator-config.ts`. Do not hand-edit between the GENERATED fences: run `bun run --cwd packages/cli generate:skill`._
317
317
 
318
318
  ### Top-level `evaluators.yaml`
319
319
 
@@ -330,10 +330,10 @@ _Generated from `EvalConfigYamlSchema` in `@eigenpal/types/src/eval/evaluator-co
330
330
  | `description` | string | no | | One-sentence explanation, in business language, of what this evaluator measures and why a stakeholder should care. Shown in the dashboard to non-technical reviewers (legal, ops, finance) who never see the YAML or judge prompt. Skip implementation detail (model names, thresholds, dot-paths); those live in `config`. |
331
331
  | `weight` | number | no | `1` | Relative importance in the weighted mean. The overall score is Σ(weightᵢ × scoreᵢ) / Σweight. Set to 0 to score-only without affecting overall. |
332
332
 
333
- - `type` — required discriminator: `exact-diff`, `llm-judge`, `custom-script`.
334
- - `config` — required object whose fields are defined by the selected evaluator type below.
333
+ - `type`: required discriminator: `exact-diff`, `llm-judge`, `custom-script`.
334
+ - `config`: required object whose fields are defined by the selected evaluator type below.
335
335
 
336
- ### `exact-diff` — JSON deep-diff against expected output
336
+ ### `exact-diff`: JSON deep-diff against expected output
337
337
 
338
338
  | Field | Type | Required | Default | Description |
339
339
  | --- | --- | --- | --- | --- |
@@ -356,7 +356,7 @@ _Generated from `EvalConfigYamlSchema` in `@eigenpal/types/src/eval/evaluator-co
356
356
  | `numericTolerance` | number | no | | Maximum absolute difference for numeric leaves at or under this path. Omitted = inherit, then historical relative epsilon. Explicit `0` is exact. |
357
357
 
358
358
 
359
- ### `llm-judge` — LLM-as-judge scoring
359
+ ### `llm-judge`: LLM-as-judge scoring
360
360
 
361
361
  | Field | Type | Required | Default | Description |
362
362
  | --- | --- | --- | --- | --- |
@@ -365,11 +365,11 @@ _Generated from `EvalConfigYamlSchema` in `@eigenpal/types/src/eval/evaluator-co
365
365
  | `mode` | `"continuous"` \| `"discrete"` | no | `"continuous"` | Continuous = the LLM returns a free-form score in [0, 1]. Discrete = the LLM picks one of your labels and the score is looked up from the table. |
366
366
  | `labels` | record<string, number> | no | | Allowed labels and the score each maps to. Required in discrete mode. The judge MUST return one of these labels; each score must be in [0, 1]. |
367
367
  | `passThreshold` | number | no | | Legacy per-evaluator pass threshold. Pass/fail now uses the single workflow-level pass threshold; this is read only to preserve the run gate of configs authored before the single-threshold model. |
368
- | `promptExtension` | string | yes | | Required. Evaluation criteria appended to the judge prompt — describe what makes a good answer for this workflow (tone, completeness, edge cases). Cannot be empty: an empty prompt yields the model inventing its own criteria. |
368
+ | `promptExtension` | string | yes | | Required. Evaluation criteria appended to the judge prompt: describe what makes a good answer for this workflow (tone, completeness, edge cases). Cannot be empty: an empty prompt yields the model inventing its own criteria. |
369
369
  | `docLabel` | string | no | `"output"` | Name the judge uses to refer to the output in its prompt (e.g. 'invoice', 'extraction'). Default is 'output'. |
370
370
 
371
371
 
372
- ### `custom-script` — JavaScript in the sandbox
372
+ ### `custom-script`: JavaScript in the sandbox
373
373
 
374
374
  | Field | Type | Required | Default | Description |
375
375
  | --- | --- | --- | --- | --- |
@@ -610,7 +610,7 @@ Generated from the same discriminated union used by evaluator validation; YAML k
610
610
  "promptExtension": {
611
611
  "type": "string",
612
612
  "minLength": 10,
613
- "description": "Required. Evaluation criteria appended to the judge prompt — describe what makes a good answer for this workflow (tone, completeness, edge cases). Cannot be empty: an empty prompt yields the model inventing its own criteria."
613
+ "description": "Required. Evaluation criteria appended to the judge prompt: describe what makes a good answer for this workflow (tone, completeness, edge cases). Cannot be empty: an empty prompt yields the model inventing its own criteria."
614
614
  },
615
615
  "docLabel": {
616
616
  "default": "output",
@@ -473,11 +473,11 @@ compose with `control.if` for legacy gating.
473
473
  <!-- GENERATED:STEP_CATALOG START -->
474
474
  ## Full catalog
475
475
 
476
- _Generated from `STEP_SCHEMAS` in `@eigenpal/types/src/workflow/step-configs.ts`. Do not hand-edit between the GENERATED fences — run `bun run --cwd packages/cli generate:skill`._
476
+ _Generated from `STEP_SCHEMAS` in `@eigenpal/types/src/workflow/step-configs.ts`. Do not hand-edit between the GENERATED fences: run `bun run --cwd packages/cli generate:skill`._
477
477
 
478
- ### AI steps — model-backed processing
478
+ ### AI steps: model-backed processing
479
479
 
480
- #### `ai.parse-v2` — Parse Document
480
+ #### `ai.parse-v2`: Parse Document
481
481
 
482
482
  Automatically read native, scanned, and mixed documents with capability-aware OCR/vision fallback and explicit completeness.
483
483
 
@@ -1724,7 +1724,7 @@ Output schema:
1724
1724
  ```
1725
1725
 
1726
1726
 
1727
- #### `ai.parse` — Parse Document v1
1727
+ #### `ai.parse`: Parse Document v1
1728
1728
 
1729
1729
  Parse Document v1, the retired parser, kept only so existing workflows run unchanged. Do not add it to workflows: use ai.parse-v2 (Parse Document), which picks native text, OCR or vision per page.
1730
1730
 
@@ -2785,7 +2785,7 @@ Output schema:
2785
2785
  ```
2786
2786
 
2787
2787
 
2788
- #### `ai.extract` — Extract Data
2788
+ #### `ai.extract`: Extract Data
2789
2789
 
2790
2790
  Extract structured data from text using AI with a JSON schema
2791
2791
 
@@ -2976,7 +2976,7 @@ Output schema:
2976
2976
  ```
2977
2977
 
2978
2978
 
2979
- #### `ai.split` — Split Document
2979
+ #### `ai.split`: Split Document
2980
2980
 
2981
2981
  Split a parsed document into named sections using an LLM. Consumes ai.parse output; emits per-section page ranges and text ready for downstream ai.extract via control.parallel_map.
2982
2982
 
@@ -3446,7 +3446,7 @@ Output schema:
3446
3446
  ```
3447
3447
 
3448
3448
 
3449
- #### `ai.segment` — Separate Documents
3449
+ #### `ai.segment`: Separate Documents
3450
3450
 
3451
3451
  Separate a concatenated batch (one big scan) into typed document instances using an LLM. Consumes ai.parse output and a type taxonomy; discovers an unknown number of documents in any order and emits per-document page ranges + text + type, ready for type-specific ai.extract via control.parallel_map. The inverse of ai.split.
3452
3452
 
@@ -3811,7 +3811,7 @@ Output schema:
3811
3811
  ```
3812
3812
 
3813
3813
 
3814
- #### `ai.classify` — Classify
3814
+ #### `ai.classify`: Classify
3815
3815
 
3816
3816
  Classify a document or text into one of a fixed label set using an LLM. Output exposes the picked label (constrained to the configured names), a coarse confidence, and a short justification. Pair with control.fail to reject documents that match an undesired label.
3817
3817
 
@@ -3951,7 +3951,7 @@ Output schema:
3951
3951
  ```
3952
3952
 
3953
3953
 
3954
- #### `ai.classify-pages` — Label Pages
3954
+ #### `ai.classify-pages`: Label Pages
3955
3955
 
3956
3956
  Assign zero or more labels to each page independently (multi-label) using an LLM. Consumes ai.parse output; emits per-page labels and a byLabel map (label -> page indices) that supports NON-contiguous selections. Feed byLabel.<label> straight into ai.vision `pageIndices` to inspect scattered pages of a type (e.g. every signature or property-photo page).
3957
3957
 
@@ -4121,7 +4121,7 @@ Output schema:
4121
4121
  ```
4122
4122
 
4123
4123
 
4124
- #### `ai.vision` — Inspect Pages (Vision)
4124
+ #### `ai.vision`: Inspect Pages (Vision)
4125
4125
 
4126
4126
  Inspect rendered page images with a vision model and return structured JSON matching a schema. The visual counterpart to Extract Data: use it for conclusions that live in the pixels rather than the text (is the document signed? are the photos usable?). Renders PDF, image, or Office/Word inputs; route to specific pages with an ai.split page range to keep it cheap.
4127
4127
 
@@ -4355,7 +4355,7 @@ Output schema:
4355
4355
  ```
4356
4356
 
4357
4357
 
4358
- #### `ai.search-files` — Search Files
4358
+ #### `ai.search-files`: Search Files
4359
4359
 
4360
4360
  Investigate a ZIP archive with a question and return an answer plus file citations. Read-only: the step opens files in the archive and never writes back. ZIP is the only accepted format in this version. Optionally scope the search to a folder prefix and cap how many files and investigation turns may run.
4361
4361
 
@@ -4505,7 +4505,7 @@ Output schema:
4505
4505
  ```
4506
4506
 
4507
4507
 
4508
- #### `ai.figure-crops` — Figure Crops
4508
+ #### `ai.figure-crops`: Figure Crops
4509
4509
 
4510
4510
  Crop and caption figure elements from a parsed PDF. Renders the source document, crops each kind=figure bbox from ai.parse output, and describes the crop with a vision model. Output is captions plus normalized bboxes for evidence viewers.
4511
4511
 
@@ -4715,9 +4715,9 @@ Output schema:
4715
4715
  ```
4716
4716
 
4717
4717
 
4718
- ### Transform steps — deterministic data transforms
4718
+ ### Transform steps: deterministic data transforms
4719
4719
 
4720
- #### `transform.set` — Set Value
4720
+ #### `transform.set`: Set Value
4721
4721
 
4722
4722
  Set key-value pairs in the output object
4723
4723
 
@@ -4785,7 +4785,7 @@ Output schema:
4785
4785
  ```
4786
4786
 
4787
4787
 
4788
- #### `transform.remove` — Remove Fields
4788
+ #### `transform.remove`: Remove Fields
4789
4789
 
4790
4790
  Remove specified fields from an object
4791
4791
 
@@ -4853,7 +4853,7 @@ Output schema:
4853
4853
  ```
4854
4854
 
4855
4855
 
4856
- #### `transform.combine` — Combine Data
4856
+ #### `transform.combine`: Combine Data
4857
4857
 
4858
4858
  Merge multiple objects or concatenate arrays
4859
4859
 
@@ -4937,7 +4937,7 @@ Output schema:
4937
4937
  ```
4938
4938
 
4939
4939
 
4940
- #### `transform.split` — Split Data
4940
+ #### `transform.split`: Split Data
4941
4941
 
4942
4942
  Split a string by delimiter or extract keys from an object
4943
4943
 
@@ -5056,7 +5056,7 @@ Output schema:
5056
5056
  ```
5057
5057
 
5058
5058
 
5059
- #### `transform.merge` — Merge Inputs
5059
+ #### `transform.merge`: Merge Inputs
5060
5060
 
5061
5061
  Merge multiple named inputs into a single output
5062
5062
 
@@ -5141,7 +5141,7 @@ Output schema:
5141
5141
  ```
5142
5142
 
5143
5143
 
5144
- #### `transform.template` — Fill Template
5144
+ #### `transform.template`: Fill Template
5145
5145
 
5146
5146
  Fill a DOCX or XLSX workspace template with data from previous steps. Reference a tmpl_... id, or a local ./templates/file.xlsx path (CLI push uploads it). Git agent templates use the fill-template skill instead and cannot be referenced here. File IDs (file_...) are not accepted. Placeholders are auto-detected. In the Office file use {placeholder} and {table:array.prop}; {{placeholder}} in an XLSX file is rejected ({{ }} is YAML Liquid only).
5147
5147
 
@@ -5241,7 +5241,7 @@ Output schema:
5241
5241
  ```
5242
5242
 
5243
5243
 
5244
- #### `transform.pdf-embed` — Embed PDF Text
5244
+ #### `transform.pdf-embed`: Embed PDF Text
5245
5245
 
5246
5246
  Embed OCR text layer into scanned PDFs/images to make them searchable
5247
5247
 
@@ -5340,7 +5340,7 @@ Output schema:
5340
5340
  ```
5341
5341
 
5342
5342
 
5343
- #### `transform.crop-regions` — Crop Regions
5343
+ #### `transform.crop-regions`: Crop Regions
5344
5344
 
5345
5345
  Render a PDF or image and crop normalized bounding boxes into JPEG run output artifacts. Pair with transform.script (bbox lists from ai.parse layout) and ai.extract or ai.vision for captions.
5346
5346
 
@@ -5602,7 +5602,7 @@ Output schema:
5602
5602
  ```
5603
5603
 
5604
5604
 
5605
- #### `transform.xlsx-to-json` — Spreadsheet to JSON
5605
+ #### `transform.xlsx-to-json`: Spreadsheet to JSON
5606
5606
 
5607
5607
  Convert an XLS or XLSX spreadsheet to a JSON array of row objects. Supports headerless files, named or positional columns, displayed text, and a single rectangular range.
5608
5608
 
@@ -5997,7 +5997,7 @@ Output schema:
5997
5997
  ```
5998
5998
 
5999
5999
 
6000
- #### `transform.json-to-xlsx` — JSON to XLSX
6000
+ #### `transform.json-to-xlsx`: JSON to XLSX
6001
6001
 
6002
6002
  Convert JSON rows into an XLSX spreadsheet. Pair with transform.xlsx-to-json. Formula-looking strings stay text; nested cell values are rejected.
6003
6003
 
@@ -6270,7 +6270,7 @@ Output schema:
6270
6270
  ```
6271
6271
 
6272
6272
 
6273
- #### `transform.script` — Script
6273
+ #### `transform.script`: Script
6274
6274
 
6275
6275
  Execute a TypeScript function in a QuickJS sandbox. Input keys become function parameters in any order, and the required `: R` return-type annotation IS this step's output schema: `inputs: { items, taxRate }` accepts `function script(taxRate: …, items: …): R { … }`.
6276
6276
 
@@ -6347,7 +6347,7 @@ Output schema:
6347
6347
  ```
6348
6348
 
6349
6349
 
6350
- #### `transform.text-chunker` — Text Chunker
6350
+ #### `transform.text-chunker`: Text Chunker
6351
6351
 
6352
6352
  Split long text into chunks with regex-anchored boundaries, overlap, and header preservation. Accepts raw text or a parsed-document object; chunks carry source page indexes when pages are provided.
6353
6353
 
@@ -6579,7 +6579,7 @@ Output schema:
6579
6579
  ```
6580
6580
 
6581
6581
 
6582
- #### `transform.regex-extract` — Regex Extract
6582
+ #### `transform.regex-extract`: Regex Extract
6583
6583
 
6584
6584
  Pull named fields from text via regex patterns (deterministic counterpart to ai.extract). Accepts raw text or a parsed-document object; matches carry `_evidence.pageIndex` when pages are provided.
6585
6585
 
@@ -6780,7 +6780,7 @@ Output schema:
6780
6780
  ```
6781
6781
 
6782
6782
 
6783
- #### `transform.archive-list` — List Archive
6783
+ #### `transform.archive-list`: List Archive
6784
6784
 
6785
6785
  List files inside a ZIP archive. Returns sorted normalized paths with compressed and uncompressed sizes. Optionally filter by folder prefix or substring, then page with offset and limit. ZIP is the only accepted format in this version. Pair with foreach or parallel_map and transform.archive-extract to open individual files.
6786
6786
 
@@ -6902,7 +6902,7 @@ Output schema:
6902
6902
  ```
6903
6903
 
6904
6904
 
6905
- #### `transform.archive-extract` — Extract Archive File
6905
+ #### `transform.archive-extract`: Extract Archive File
6906
6906
 
6907
6907
  Extract exactly one file from a ZIP archive and store it as a run output file. Downstream steps can pass {{ steps.extract.output }} into Parse, Vision, or another archive step. Nested ZIPs are stored as ZIP files, not unpacked. Each extraction is capped at 64 MiB. ZIP is the only accepted format in this version.
6908
6908
 
@@ -6998,7 +6998,7 @@ Output schema:
6998
6998
  ```
6999
6999
 
7000
7000
 
7001
- #### `transform.merge-documents` — Merge Documents
7001
+ #### `transform.merge-documents`: Merge Documents
7002
7002
 
7003
7003
  Concatenate several files into one PDF run output, in order. Accepts PDFs, images (multi-page TIFF included) and Word, PowerPoint, OpenDocument or RTF documents. Use it when one logical document arrives as several uploads: pass {{ steps.merge.output.file }} into Parse so page labels, page ranges and vision page indices stay continuous across every source. Output sources[] maps each input file to its merged page range.
7004
7004
 
@@ -7152,9 +7152,9 @@ Output schema:
7152
7152
  ```
7153
7153
 
7154
7154
 
7155
- ### Action steps — external side effects
7155
+ ### Action steps: external side effects
7156
7156
 
7157
- #### `action.http` — HTTP Request
7157
+ #### `action.http`: HTTP Request
7158
7158
 
7159
7159
  Make an HTTP request to an external API
7160
7160
 
@@ -7285,7 +7285,7 @@ Output schema:
7285
7285
  ```
7286
7286
 
7287
7287
 
7288
- #### `action.invoke-workflow` — Invoke Workflow
7288
+ #### `action.invoke-workflow`: Invoke Workflow
7289
7289
 
7290
7290
  Execute another workflow and return its output
7291
7291
 
@@ -7395,7 +7395,7 @@ Output schema:
7395
7395
  ```
7396
7396
 
7397
7397
 
7398
- #### `action.website-reader` — Website Reader
7398
+ #### `action.website-reader`: Website Reader
7399
7399
 
7400
7400
  Fetch a webpage and convert content to markdown
7401
7401
 
@@ -7527,13 +7527,13 @@ Output schema:
7527
7527
  ```
7528
7528
 
7529
7529
 
7530
- #### `action.email` — Send Email
7530
+ #### `action.email`: Send Email
7531
7531
 
7532
7532
  Send an email with optional attachments via a selected email server. At most 10 recipients combined across to, cc, and bcc.
7533
7533
 
7534
7534
  **Behavior and examples:** `eigenpal docs read steps/action/email`
7535
7535
 
7536
- **Durable retry:** This control step is not retried durably.
7536
+ **Durable retry:** Supported only for failures where the provider did not accept the message: SMTP connection errors before the message is sent, and Resend network errors, rate limits and server failures (Resend dedupes on an idempotency key). A message that may have been accepted is never resent.
7537
7537
 
7538
7538
  **Config** (in `step.with`):
7539
7539
 
@@ -7769,9 +7769,9 @@ Output schema:
7769
7769
  ```
7770
7770
 
7771
7771
 
7772
- ### Control steps — flow control
7772
+ ### Control steps: flow control
7773
7773
 
7774
- #### `control.if` — Condition
7774
+ #### `control.if`: Condition
7775
7775
 
7776
7776
  Branch execution based on a condition expression
7777
7777
 
@@ -7849,7 +7849,7 @@ Output schema:
7849
7849
  ```
7850
7850
 
7851
7851
 
7852
- #### `control.switch` — Switch
7852
+ #### `control.switch`: Switch
7853
7853
 
7854
7854
  Multi-way routing: resolve an expression and run the first case whose value matches (else default). Cleaner than a nested control.if chain for routing an item to one of N pipelines by a discriminator field like a document type.
7855
7855
 
@@ -7957,7 +7957,7 @@ Output schema:
7957
7957
  ```
7958
7958
 
7959
7959
 
7960
- #### `control.foreach` — For Each
7960
+ #### `control.foreach`: For Each
7961
7961
 
7962
7962
  Loop over an array and execute steps for each item
7963
7963
 
@@ -8045,7 +8045,7 @@ Output schema:
8045
8045
  ```
8046
8046
 
8047
8047
 
8048
- #### `control.parallel_map` — Parallel Map
8048
+ #### `control.parallel_map`: Parallel Map
8049
8049
 
8050
8050
  Iterate over an array with concurrent execution up to a limit
8051
8051
 
@@ -8141,7 +8141,7 @@ Output schema:
8141
8141
  ```
8142
8142
 
8143
8143
 
8144
- #### `control.parallel` — Parallel
8144
+ #### `control.parallel`: Parallel
8145
8145
 
8146
8146
  Execute multiple branches concurrently
8147
8147
 
@@ -8186,7 +8186,7 @@ Output schema:
8186
8186
  ```
8187
8187
 
8188
8188
 
8189
- #### `control.wait` — Wait
8189
+ #### `control.wait`: Wait
8190
8190
 
8191
8191
  Pause workflow execution for a specified duration
8192
8192
 
@@ -8249,7 +8249,7 @@ Output schema:
8249
8249
  ```
8250
8250
 
8251
8251
 
8252
- #### `control.human_review` — Human Review
8252
+ #### `control.human_review`: Human Review
8253
8253
 
8254
8254
  Pause the run for selective field confirmation or correction, then continue with the reviewed data.
8255
8255
 
@@ -8695,7 +8695,7 @@ Output schema:
8695
8695
  ```
8696
8696
 
8697
8697
 
8698
- #### `control.fail` — Fail
8698
+ #### `control.fail`: Fail
8699
8699
 
8700
8700
  Terminate the workflow with a typed status code + message. With an optional condition, only fails when the condition is truthy; otherwise always fails when reached. Pair with ai.classify or any prior step to fail fast on bad inputs.
8701
8701