From 982f708c85cb998705be5c6a364c034becc06bc6 Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Wed, 21 Aug 2024 09:23:51 -0700
Subject: [PATCH 01/10] Improve ListArtifact

---
 CHANGELOG.md                               |  2 ++
 griptape/artifacts/list_artifact.py        | 23 +++++++++++++---------
 griptape/schemas/base_schema.py            |  6 +++++-
 griptape/tasks/tool_task.py                |  6 +++++-
 tests/unit/artifacts/test_list_artifact.py |  6 ++++++
 5 files changed, 32 insertions(+), 11 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 0528b1a46..0aedd1970 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,6 +7,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## Unreleased
 ### Added
 - `BaseConversationMemory.prompt_driver` for use with autopruning. 
+- Generic type support to `ListArtifact`.
+- Iteration support to `ListArtifact`.
 
 ### Fixed
 - Parsing streaming response with some OpenAi compatible services.
diff --git a/griptape/artifacts/list_artifact.py b/griptape/artifacts/list_artifact.py
index 298f29c6a..9ebec394c 100644
--- a/griptape/artifacts/list_artifact.py
+++ b/griptape/artifacts/list_artifact.py
@@ -1,23 +1,25 @@
 from __future__ import annotations
 
-from typing import TYPE_CHECKING, Optional
+from typing import TYPE_CHECKING, Generic, Optional, TypeVar
 
 from attrs import Attribute, define, field
 
 from griptape.artifacts import BaseArtifact
 
 if TYPE_CHECKING:
-    from collections.abc import Sequence
+    from collections.abc import Iterator, Sequence
+
+T = TypeVar("T", bound=BaseArtifact)
 
 
 @define
-class ListArtifact(BaseArtifact):
-    value: Sequence[BaseArtifact] = field(factory=list, metadata={"serializable": True})
+class ListArtifact(BaseArtifact, Generic[T]):
+    value: Sequence[T] = field(factory=list, metadata={"serializable": True})
     item_separator: str = field(default="\n\n", kw_only=True, metadata={"serializable": True})
     validate_uniform_types: bool = field(default=False, kw_only=True, metadata={"serializable": True})
 
     @value.validator  # pyright: ignore[reportAttributeAccessIssue]
-    def validate_value(self, _: Attribute, value: list[BaseArtifact]) -> None:
+    def validate_value(self, _: Attribute, value: list[T]) -> None:
         if self.validate_uniform_types and len(value) > 0:
             first_type = type(value[0])
 
@@ -31,18 +33,21 @@ def child_type(self) -> Optional[type]:
         else:
             return None
 
-    def __getitem__(self, key: int) -> BaseArtifact:
+    def __getitem__(self, key: int) -> T:
         return self.value[key]
 
     def __bool__(self) -> bool:
         return len(self) > 0
 
+    def __add__(self, other: BaseArtifact) -> ListArtifact[T]:
+        return ListArtifact(self.value + other.value)
+
+    def __iter__(self) -> Iterator[T]:
+        return iter(self.value)
+
     def to_text(self) -> str:
         return self.item_separator.join([v.to_text() for v in self.value])
 
-    def __add__(self, other: BaseArtifact) -> BaseArtifact:
-        return ListArtifact(self.value + other.value)
-
     def is_type(self, target_type: type) -> bool:
         if self.value:
             return isinstance(self.value[0], target_type)
diff --git a/griptape/schemas/base_schema.py b/griptape/schemas/base_schema.py
index f25e8870b..b285d1476 100644
--- a/griptape/schemas/base_schema.py
+++ b/griptape/schemas/base_schema.py
@@ -2,7 +2,7 @@
 
 from abc import ABC
 from collections.abc import Sequence
-from typing import Any, Literal, Union, _SpecialForm, get_args, get_origin
+from typing import Any, Literal, TypeVar, Union, _SpecialForm, get_args, get_origin
 
 import attrs
 from marshmallow import INCLUDE, Schema, fields
@@ -56,6 +56,10 @@ def _get_field_for_type(cls, field_type: type) -> fields.Field | fields.Nested:
 
         field_class, args, optional = cls._get_field_type_info(field_type)
 
+        # Resolve TypeVars to their bound type
+        if isinstance(field_class, TypeVar):
+            field_class = field_class.__bound__
+
         if attrs.has(field_class):
             if ABC in field_class.__bases__:
                 return fields.Nested(PolymorphicSchema(inner_class=field_class), allow_none=optional)
diff --git a/griptape/tasks/tool_task.py b/griptape/tasks/tool_task.py
index 6dd5000b3..68260ea91 100644
--- a/griptape/tasks/tool_task.py
+++ b/griptape/tasks/tool_task.py
@@ -84,7 +84,11 @@ def run(self) -> BaseArtifact:
             subtask.after_run()
 
             if isinstance(subtask.output, ListArtifact):
-                self.output = subtask.output[0]
+                first_artifact = subtask.output[0]
+                if isinstance(first_artifact, BaseArtifact):
+                    self.output = first_artifact
+                else:
+                    self.output = ErrorArtifact(f"Output is not an Artifact: {type(subtask.output[0])}")
             else:
                 self.output = InfoArtifact("No tool output")
         except Exception as e:
diff --git a/tests/unit/artifacts/test_list_artifact.py b/tests/unit/artifacts/test_list_artifact.py
index 06d234645..37769e1b8 100644
--- a/tests/unit/artifacts/test_list_artifact.py
+++ b/tests/unit/artifacts/test_list_artifact.py
@@ -23,6 +23,12 @@ def test___add__(self):
         assert artifact.value[0].value == "foo"
         assert artifact.value[1].value == "bar"
 
+    def test___iter__(self):
+        assert [a.value for a in ListArtifact([TextArtifact("foo"), TextArtifact("bar")])] == ["foo", "bar"]
+
+    def test_type_var(self):
+        assert ListArtifact[TextArtifact]([TextArtifact("foo")]).value[0].value == "foo"
+
     def test_validate_value(self):
         with pytest.raises(ValueError):
             ListArtifact([TextArtifact("foo"), BlobArtifact(b"bar")], validate_uniform_types=True)

From b839973e8027c0fb075b6d6821bc4966699ab72a Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Wed, 21 Aug 2024 13:28:29 -0700
Subject: [PATCH 02/10] Refactor extraction engine methods, return type, and
 extraction behavior.

---
 CHANGELOG.md                                  |  7 ++++
 .../engines/src/extraction_engines_1.py       | 12 +++----
 .../engines/src/extraction_engines_2.py       | 17 ++++------
 .../extraction/base_extraction_engine.py      | 17 +++++++---
 .../extraction/csv_extraction_engine.py       | 26 +++++++--------
 .../extraction/json_extraction_engine.py      | 33 ++++++++-----------
 griptape/tasks/extraction_task.py             |  7 ++--
 .../engines/extraction/json/system.j2         |  2 ++
 .../templates/engines/extraction/json/user.j2 |  4 +--
 griptape/tools/extraction/tool.py             |  2 +-
 .../extraction/test_csv_extraction_engine.py  |  4 +--
 .../extraction/test_json_extraction_engine.py | 29 +++++++++-------
 tests/unit/memory/tool/test_task_memory.py    |  6 +---
 tests/unit/tools/test_extraction_tool.py      | 10 +++---
 14 files changed, 91 insertions(+), 85 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 0aedd1970..09e82ef68 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -10,6 +10,13 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Generic type support to `ListArtifact`.
 - Iteration support to `ListArtifact`.
 
+### Changed
+- **BREAKING**: Split `BaseExtractionEngine.extract` into `extract` and `extract_artifacts` for consistency with `BaseSummaryEngine`.
+- **BREAKING**: `BaseExtractionEngine` no longer catches exceptions and returns `ErrorArtifact`s.
+- `JsonExtractionEngine` to extract either a JSON object or array depending on the provided schema. 
+- `JsonExtractionEngine.extract_artifacts` now returns a `ListArtifact[JsonArtifact]`.
+- `CsvExtractionEngine.extract_artifacts` now returns a `ListArtifact[CsvRowArtifact]`.
+
 ### Fixed
 - Parsing streaming response with some OpenAi compatible services.
 
diff --git a/docs/griptape-framework/engines/src/extraction_engines_1.py b/docs/griptape-framework/engines/src/extraction_engines_1.py
index c681980f2..45ccfd3e0 100644
--- a/docs/griptape-framework/engines/src/extraction_engines_1.py
+++ b/docs/griptape-framework/engines/src/extraction_engines_1.py
@@ -1,10 +1,9 @@
-from griptape.artifacts import ListArtifact
 from griptape.drivers import OpenAiChatPromptDriver
 from griptape.engines import CsvExtractionEngine
 
 # Initialize the CsvExtractionEngine instance
 csv_engine = CsvExtractionEngine(
-    prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"),
+    prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"), column_names=["name", "age", "location"]
 )
 
 # Define some unstructured data
@@ -15,10 +14,7 @@
 """
 
 # Extract CSV rows using the engine
-result = csv_engine.extract(sample_text, column_names=["name", "age", "location"])
+result = csv_engine.extract_text(sample_text)
 
-if isinstance(result, ListArtifact):
-    for row in result.value:
-        print(row.to_text())
-else:
-    print(result.to_text())
+for row in result:
+    print(row.to_text())
diff --git a/docs/griptape-framework/engines/src/extraction_engines_2.py b/docs/griptape-framework/engines/src/extraction_engines_2.py
index d47bb48e5..2fb8cd8b0 100644
--- a/docs/griptape-framework/engines/src/extraction_engines_2.py
+++ b/docs/griptape-framework/engines/src/extraction_engines_2.py
@@ -1,11 +1,13 @@
 from schema import Schema
 
-from griptape.artifacts.list_artifact import ListArtifact
 from griptape.drivers import OpenAiChatPromptDriver
 from griptape.engines import JsonExtractionEngine
 
+# Define a schema for extraction
+user_schema = Schema([{"name": str, "age": int, "location": str}]).json_schema("UserSchema")
+
 json_engine = JsonExtractionEngine(
-    prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"),
+    prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"), template_schema=user_schema
 )
 
 # Define some unstructured data
@@ -14,14 +16,9 @@
 Bob (Age 35) lives in California.
 """
 
-# Define a schema for extraction
-user_schema = Schema({"users": [{"name": str, "age": int, "location": str}]}).json_schema("UserSchema")
 
 # Extract data using the engine
-result = json_engine.extract(sample_json_text, template_schema=user_schema)
+result = json_engine.extract_text(sample_json_text)
 
-if isinstance(result, ListArtifact):
-    for artifact in result.value:
-        print(artifact.value)
-else:
-    print(result.to_text())
+for artifact in result:
+    print(artifact.value)
diff --git a/griptape/engines/extraction/base_extraction_engine.py b/griptape/engines/extraction/base_extraction_engine.py
index fb1fab6c4..10ba5d142 100644
--- a/griptape/engines/extraction/base_extraction_engine.py
+++ b/griptape/engines/extraction/base_extraction_engine.py
@@ -5,11 +5,11 @@
 
 from attrs import Attribute, Factory, define, field
 
+from griptape.artifacts import ListArtifact, TextArtifact
 from griptape.chunkers import BaseChunker, TextChunker
 from griptape.configs import Defaults
 
 if TYPE_CHECKING:
-    from griptape.artifacts import ErrorArtifact, ListArtifact
     from griptape.drivers import BasePromptDriver
     from griptape.rules import Ruleset
 
@@ -47,11 +47,20 @@ def min_response_tokens(self) -> int:
             - self.prompt_driver.tokenizer.max_input_tokens * self.max_token_multiplier,
         )
 
+    def extract_text(
+        self,
+        text: str,
+        *,
+        rulesets: Optional[list[Ruleset]] = None,
+        **kwargs,
+    ) -> ListArtifact:
+        return self.extract_artifacts(ListArtifact([TextArtifact(text)]), rulesets=rulesets, **kwargs)
+
     @abstractmethod
-    def extract(
+    def extract_artifacts(
         self,
-        text: str | ListArtifact,
+        artifacts: ListArtifact[TextArtifact],
         *,
         rulesets: Optional[list[Ruleset]] = None,
         **kwargs,
-    ) -> ListArtifact | ErrorArtifact: ...
+    ) -> ListArtifact: ...
diff --git a/griptape/engines/extraction/csv_extraction_engine.py b/griptape/engines/extraction/csv_extraction_engine.py
index c9c040f65..40c9058ec 100644
--- a/griptape/engines/extraction/csv_extraction_engine.py
+++ b/griptape/engines/extraction/csv_extraction_engine.py
@@ -6,7 +6,7 @@
 
 from attrs import Factory, define, field
 
-from griptape.artifacts import CsvRowArtifact, ErrorArtifact, ListArtifact, TextArtifact
+from griptape.artifacts import CsvRowArtifact, ListArtifact, TextArtifact
 from griptape.common import Message, PromptStack
 from griptape.engines import BaseExtractionEngine
 from griptape.utils import J2
@@ -21,23 +21,21 @@ class CsvExtractionEngine(BaseExtractionEngine):
     system_template_generator: J2 = field(default=Factory(lambda: J2("engines/extraction/csv/system.j2")), kw_only=True)
     user_template_generator: J2 = field(default=Factory(lambda: J2("engines/extraction/csv/user.j2")), kw_only=True)
 
-    def extract(
+    def extract_artifacts(
         self,
-        text: str | ListArtifact,
+        artifacts: ListArtifact[TextArtifact],
         *,
         rulesets: Optional[list[Ruleset]] = None,
         **kwargs,
-    ) -> ListArtifact | ErrorArtifact:
-        try:
-            return ListArtifact(
-                self._extract_rec(
-                    cast(list[TextArtifact], text.value) if isinstance(text, ListArtifact) else [TextArtifact(text)],
-                    [],
-                ),
-                item_separator="\n",
-            )
-        except Exception as e:
-            return ErrorArtifact(f"error extracting CSV rows: {e}")
+    ) -> ListArtifact[CsvRowArtifact]:
+        return ListArtifact(
+            self._extract_rec(
+                cast(list[TextArtifact], artifacts.value),
+                [],
+                rulesets=rulesets,
+            ),
+            item_separator="\n",
+        )
 
     def text_to_csv_rows(self, text: str, column_names: list[str]) -> list[CsvRowArtifact]:
         rows = []
diff --git a/griptape/engines/extraction/json_extraction_engine.py b/griptape/engines/extraction/json_extraction_engine.py
index 8f2f4a3fe..d3928c3ea 100644
--- a/griptape/engines/extraction/json_extraction_engine.py
+++ b/griptape/engines/extraction/json_extraction_engine.py
@@ -6,7 +6,7 @@
 
 from attrs import Factory, define, field
 
-from griptape.artifacts import ErrorArtifact, ListArtifact, TextArtifact
+from griptape.artifacts import JsonArtifact, ListArtifact, TextArtifact
 from griptape.common import PromptStack
 from griptape.common.prompt_stack.messages.message import Message
 from griptape.engines import BaseExtractionEngine
@@ -18,7 +18,7 @@
 
 @define
 class JsonExtractionEngine(BaseExtractionEngine):
-    JSON_PATTERN = r"(?s)[^\[]*(\[.*\])"
+    JSON_PATTERN = r"(?s)(\{.*\}|\[.*\])"
 
     template_schema: dict = field(default=Factory(dict), kw_only=True)
     system_template_generator: J2 = field(
@@ -26,40 +26,33 @@ class JsonExtractionEngine(BaseExtractionEngine):
     )
     user_template_generator: J2 = field(default=Factory(lambda: J2("engines/extraction/json/user.j2")), kw_only=True)
 
-    def extract(
+    def extract_artifacts(
         self,
-        text: str | ListArtifact,
+        artifacts: ListArtifact[TextArtifact],
         *,
         rulesets: Optional[list[Ruleset]] = None,
         **kwargs,
-    ) -> ListArtifact | ErrorArtifact:
-        try:
-            return ListArtifact(
-                self._extract_rec(
-                    cast(list[TextArtifact], text.value) if isinstance(text, ListArtifact) else [TextArtifact(text)],
-                    [],
-                    rulesets=rulesets,
-                ),
-                item_separator="\n",
-            )
-        except Exception as e:
-            return ErrorArtifact(f"error extracting JSON: {e}")
+    ) -> ListArtifact[JsonArtifact]:
+        return ListArtifact(
+            self._extract_rec(cast(list[TextArtifact], artifacts.value), [], rulesets=rulesets),
+            item_separator="\n",
+        )
 
-    def json_to_text_artifacts(self, json_input: str) -> list[TextArtifact]:
+    def json_to_text_artifacts(self, json_input: str) -> list[JsonArtifact]:
         json_matches = re.findall(self.JSON_PATTERN, json_input, re.DOTALL)
 
         if json_matches:
-            return [TextArtifact(json.dumps(e)) for e in json.loads(json_matches[-1])]
+            return [JsonArtifact(json.loads(e)) for e in json_matches]
         else:
             return []
 
     def _extract_rec(
         self,
         artifacts: list[TextArtifact],
-        extractions: list[TextArtifact],
+        extractions: list[JsonArtifact],
         *,
         rulesets: Optional[list[Ruleset]] = None,
-    ) -> list[TextArtifact]:
+    ) -> list[JsonArtifact]:
         artifacts_text = self.chunk_joiner.join([a.value for a in artifacts])
         system_prompt = self.system_template_generator.render(
             json_template_schema=json.dumps(self.template_schema),
diff --git a/griptape/tasks/extraction_task.py b/griptape/tasks/extraction_task.py
index c74c3ac49..234840401 100644
--- a/griptape/tasks/extraction_task.py
+++ b/griptape/tasks/extraction_task.py
@@ -4,10 +4,11 @@
 
 from attrs import define, field
 
+from griptape.artifacts import ListArtifact
 from griptape.tasks import BaseTextInputTask
 
 if TYPE_CHECKING:
-    from griptape.artifacts import ErrorArtifact, ListArtifact
+    from griptape.artifacts import ErrorArtifact
     from griptape.engines import BaseExtractionEngine
 
 
@@ -17,4 +18,6 @@ class ExtractionTask(BaseTextInputTask):
     args: dict = field(kw_only=True, factory=dict)
 
     def run(self) -> ListArtifact | ErrorArtifact:
-        return self.extraction_engine.extract(self.input.to_text(), rulesets=self.all_rulesets, **self.args)
+        return self.extraction_engine.extract_artifacts(
+            ListArtifact([self.input]), rulesets=self.all_rulesets, **self.args
+        )
diff --git a/griptape/templates/engines/extraction/json/system.j2 b/griptape/templates/engines/extraction/json/system.j2
index 987ff19a9..568914dd5 100644
--- a/griptape/templates/engines/extraction/json/system.j2
+++ b/griptape/templates/engines/extraction/json/system.j2
@@ -1,3 +1,5 @@
+Your output must exactly match the Extraction Template JSON Schema.
+
 Extraction Template JSON Schema: """{{ json_template_schema }}"""
 
 {% if rulesets %}
diff --git a/griptape/templates/engines/extraction/json/user.j2 b/griptape/templates/engines/extraction/json/user.j2
index 984977d9a..9b0a611f0 100644
--- a/griptape/templates/engines/extraction/json/user.j2
+++ b/griptape/templates/engines/extraction/json/user.j2
@@ -1,4 +1,4 @@
-Extract information from the Text based on the Extraction Template JSON Schema into an array of JSON objects.
+Extract information from the Text.
 Text: """{{ text }}"""
 
-JSON array:
+JSON:
diff --git a/griptape/tools/extraction/tool.py b/griptape/tools/extraction/tool.py
index 1f6d06b80..279ca3376 100644
--- a/griptape/tools/extraction/tool.py
+++ b/griptape/tools/extraction/tool.py
@@ -57,4 +57,4 @@ def extract(self, params: dict) -> ListArtifact | InfoArtifact | ErrorArtifact:
             else:
                 return ErrorArtifact("memory not found")
 
-        return self.extraction_engine.extract(artifacts)
+        return self.extraction_engine.extract_artifacts(artifacts)
diff --git a/tests/unit/engines/extraction/test_csv_extraction_engine.py b/tests/unit/engines/extraction/test_csv_extraction_engine.py
index 893c21d60..36f70c6ec 100644
--- a/tests/unit/engines/extraction/test_csv_extraction_engine.py
+++ b/tests/unit/engines/extraction/test_csv_extraction_engine.py
@@ -8,8 +8,8 @@ class TestCsvExtractionEngine:
     def engine(self):
         return CsvExtractionEngine(column_names=["test1"])
 
-    def test_extract(self, engine):
-        result = engine.extract("foo")
+    def test_extract_text(self, engine):
+        result = engine.extract_text("foo")
 
         assert len(result.value) == 1
         assert result.value[0].value == {"test1": "mock output"}
diff --git a/tests/unit/engines/extraction/test_json_extraction_engine.py b/tests/unit/engines/extraction/test_json_extraction_engine.py
index 48430f1e5..938438af4 100644
--- a/tests/unit/engines/extraction/test_json_extraction_engine.py
+++ b/tests/unit/engines/extraction/test_json_extraction_engine.py
@@ -1,7 +1,9 @@
+from os.path import dirname, join, normpath
+from pathlib import Path
+
 import pytest
 from schema import Schema
 
-from griptape.artifacts import ErrorArtifact
 from griptape.engines import JsonExtractionEngine
 from tests.mocks.mock_prompt_driver import MockPromptDriver
 
@@ -16,22 +18,27 @@ def engine(self):
             template_schema=Schema({"foo": "bar"}).json_schema("TemplateSchema"),
         )
 
-    def test_extract(self, engine):
-        result = engine.extract("foo")
+    def test_extract_text(self, engine):
+        result = engine.extract_text("foo")
+
+        assert len(result.value) == 1
+        assert result.value[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
+
+    def test_chunked_extract_text(self, engine):
+        large_text = Path(normpath(join(dirname(__file__), "../../../resources", "test.txt"))).read_text()
 
-        assert len(result.value) == 2
-        assert result.value[0].value == '{"test_key_1": "test_value_1"}'
-        assert result.value[1].value == '{"test_key_2": "test_value_2"}'
+        extracted = engine.extract_text(large_text * 50)
+        assert len(extracted) == 177
+        assert extracted[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
 
     def test_extract_error(self, engine):
         engine.template_schema = lambda: "non serializable"
-        assert isinstance(engine.extract("foo"), ErrorArtifact)
+        with pytest.raises(TypeError):
+            engine.extract_text("foo")
 
     def test_json_to_text_artifacts(self, engine):
-        assert [
-            a.value
-            for a in engine.json_to_text_artifacts('[{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]')
-        ] == ['{"test_key_1": "test_value_1"}', '{"test_key_2": "test_value_2"}']
+        extracted = engine.json_to_text_artifacts('[{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]')
+        assert extracted[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
 
     def test_json_to_text_artifacts_no_matches(self, engine):
         assert engine.json_to_text_artifacts("asdfasdfasdf") == []
diff --git a/tests/unit/memory/tool/test_task_memory.py b/tests/unit/memory/tool/test_task_memory.py
index 2f6ffe1c9..d2575959a 100644
--- a/tests/unit/memory/tool/test_task_memory.py
+++ b/tests/unit/memory/tool/test_task_memory.py
@@ -1,6 +1,6 @@
 import pytest
 
-from griptape.artifacts import BlobArtifact, CsvRowArtifact, ErrorArtifact, InfoArtifact, ListArtifact, TextArtifact
+from griptape.artifacts import BlobArtifact, ErrorArtifact, InfoArtifact, ListArtifact, TextArtifact
 from griptape.memory import TaskMemory
 from griptape.memory.task.storage import BlobArtifactStorage, TextArtifactStorage
 from griptape.structures import Agent
@@ -10,10 +10,6 @@
 
 
 class TestTaskMemory:
-    @pytest.fixture(autouse=True)
-    def _mock_griptape(self, mocker):
-        mocker.patch("griptape.engines.CsvExtractionEngine.extract", return_value=[CsvRowArtifact({"foo": "bar"})])
-
     @pytest.fixture()
     def memory(self):
         return defaults.text_task_memory("MyMemory")
diff --git a/tests/unit/tools/test_extraction_tool.py b/tests/unit/tools/test_extraction_tool.py
index 1219da373..edf1663d2 100644
--- a/tests/unit/tools/test_extraction_tool.py
+++ b/tests/unit/tools/test_extraction_tool.py
@@ -39,16 +39,14 @@ def test_json_extract_artifacts(self, json_tool):
             {"values": {"data": {"memory_name": json_tool.input_memory[0].name, "artifact_namespace": "foo"}}}
         )
 
-        assert len(result.value) == 2
-        assert result.value[0].value == '{"test_key_1": "test_value_1"}'
-        assert result.value[1].value == '{"test_key_2": "test_value_2"}'
+        assert len(result.value) == 1
+        assert result.value[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
 
     def test_json_extract_content(self, json_tool):
         result = json_tool.extract({"values": {"data": "foo"}})
 
-        assert len(result.value) == 2
-        assert result.value[0].value == '{"test_key_1": "test_value_1"}'
-        assert result.value[1].value == '{"test_key_2": "test_value_2"}'
+        assert len(result.value) == 1
+        assert result.value[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
 
     def test_csv_extract_artifacts(self, csv_tool):
         csv_tool.input_memory[0].store_artifact("foo", TextArtifact("foo,bar\nbaz,maz"))

From e92e0850502fb7aacb12c7d72bc525bea82a7115 Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Wed, 21 Aug 2024 13:37:35 -0700
Subject: [PATCH 03/10] Update system template

---
 griptape/templates/engines/extraction/json/system.j2 | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/griptape/templates/engines/extraction/json/system.j2 b/griptape/templates/engines/extraction/json/system.j2
index 568914dd5..b6bac028a 100644
--- a/griptape/templates/engines/extraction/json/system.j2
+++ b/griptape/templates/engines/extraction/json/system.j2
@@ -1,4 +1,4 @@
-Your output must exactly match the Extraction Template JSON Schema.
+Your answer MUST be JSON that successfully validates against the Extraction Template JSON Schema.
 
 Extraction Template JSON Schema: """{{ json_template_schema }}"""
 

From 004e36a427d6e0e03a203ca48fc629a4cca3133c Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Wed, 21 Aug 2024 14:47:51 -0700
Subject: [PATCH 04/10] Revert some sillyness, update docs

---
 CHANGELOG.md                                  |  3 +-
 .../engines/extraction-engines.md             | 27 +++++++++++------
 .../engines/src/extraction_engines_2.py       | 29 ++++++++++++-------
 .../structures/src/tasks_6.py                 |  7 ++---
 .../structures/src/tasks_7.py                 |  4 +--
 griptape/artifacts/json_artifact.py           |  4 +--
 .../extraction/csv_extraction_engine.py       |  2 +-
 .../extraction/json_extraction_engine.py      |  6 ++--
 .../engines/extraction/json/system.j2         |  2 --
 .../templates/engines/extraction/json/user.j2 |  4 +--
 .../extraction/test_json_extraction_engine.py | 15 ++++++----
 tests/unit/tools/test_extraction_tool.py      | 10 ++++---
 12 files changed, 66 insertions(+), 47 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 09e82ef68..0b9ccbe59 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -13,7 +13,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Changed
 - **BREAKING**: Split `BaseExtractionEngine.extract` into `extract` and `extract_artifacts` for consistency with `BaseSummaryEngine`.
 - **BREAKING**: `BaseExtractionEngine` no longer catches exceptions and returns `ErrorArtifact`s.
-- `JsonExtractionEngine` to extract either a JSON object or array depending on the provided schema. 
+- **BREAKING**: `JsonExtractionEngine.template_schema` is now required.
+- **BREAKING**: `CsvExtractionEngine.column_names` is now required.
 - `JsonExtractionEngine.extract_artifacts` now returns a `ListArtifact[JsonArtifact]`.
 - `CsvExtractionEngine.extract_artifacts` now returns a `ListArtifact[CsvRowArtifact]`.
 
diff --git a/docs/griptape-framework/engines/extraction-engines.md b/docs/griptape-framework/engines/extraction-engines.md
index b971e63cc..43333e4b0 100644
--- a/docs/griptape-framework/engines/extraction-engines.md
+++ b/docs/griptape-framework/engines/extraction-engines.md
@@ -10,10 +10,7 @@ As of now, Griptape supports two types of Extraction Engines: the CSV Extraction
 
 ## CSV
 
-The CSV Extraction Engine is designed specifically for extracting data from CSV-formatted content.
-
-!!! info
-    The CSV Extraction Engine requires the `column_names` parameter for specifying the columns to be extracted.
+The CSV Extraction Engine is designed for extracting CSV-formatted content from unstructured data.
 
 ```python
 --8<-- "docs/griptape-framework/engines/src/extraction_engines_1.py"
@@ -27,15 +24,27 @@ Charlie,40,Texas
 
 ## JSON
 
-The JSON Extraction Engine is tailored for extracting data from JSON-formatted content. 
+The JSON Extraction Engine is designed for extracting JSON-formatted content from unstructed data. 
 
-!!! info
-    The JSON Extraction Engine requires the `template_schema` parameter for specifying the structure to be extracted.
 
 ```python
 --8<-- "docs/griptape-framework/engines/src/extraction_engines_2.py"
 ```
 ```
-{'name': 'Alice', 'age': 28, 'location': 'New York'}
-{'name': 'Bob', 'age': 35, 'location': 'California'}
+{
+  "model": "GPT-3.5",
+  "notes": [
+    "Part of OpenAI's GPT series.",
+    "Used in ChatGPT and Microsoft Copilot."
+  ]
+}
+{
+  "model": "GPT-4",
+  "notes": [
+    "Part of OpenAI's GPT series.",
+    "Praised for increased accuracy and multimodal capabilities.",
+    "Architecture and number of parameters not revealed."
+  ]
+}
+...Output truncated for brevity...
 ```
diff --git a/docs/griptape-framework/engines/src/extraction_engines_2.py b/docs/griptape-framework/engines/src/extraction_engines_2.py
index 2fb8cd8b0..35bbe53cd 100644
--- a/docs/griptape-framework/engines/src/extraction_engines_2.py
+++ b/docs/griptape-framework/engines/src/extraction_engines_2.py
@@ -1,24 +1,31 @@
-from schema import Schema
+import json
 
+from schema import Literal, Schema
+
+from griptape.artifacts import ErrorArtifact, ListArtifact
 from griptape.drivers import OpenAiChatPromptDriver
 from griptape.engines import JsonExtractionEngine
+from griptape.loaders import WebLoader
 
 # Define a schema for extraction
-user_schema = Schema([{"name": str, "age": int, "location": str}]).json_schema("UserSchema")
-
 json_engine = JsonExtractionEngine(
-    prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"), template_schema=user_schema
+    prompt_driver=OpenAiChatPromptDriver(model="gpt-4o"),
+    template_schema=Schema(
+        {
+            Literal("model", description="Name of an LLM model."): str,
+            Literal("notes", description="Any notes of substance about the model."): Schema([str]),
+        }
+    ).json_schema("ProductSchema"),
 )
 
-# Define some unstructured data
-sample_json_text = """
-Alice (Age 28) lives in New York.
-Bob (Age 35) lives in California.
-"""
+# Load data from the web
+web_data = WebLoader().load("https://en.wikipedia.org/wiki/Large_language_model")
 
+if isinstance(web_data, ErrorArtifact):
+    raise Exception(web_data.value)
 
 # Extract data using the engine
-result = json_engine.extract_text(sample_json_text)
+result = json_engine.extract_artifacts(ListArtifact(web_data))
 
 for artifact in result:
-    print(artifact.value)
+    print(json.dumps(artifact.value, indent=2))
diff --git a/docs/griptape-framework/structures/src/tasks_6.py b/docs/griptape-framework/structures/src/tasks_6.py
index a1b84e44d..ecd6f354f 100644
--- a/docs/griptape-framework/structures/src/tasks_6.py
+++ b/docs/griptape-framework/structures/src/tasks_6.py
@@ -4,7 +4,9 @@
 from griptape.tasks import ExtractionTask
 
 # Instantiate the CSV extraction engine
-csv_extraction_engine = CsvExtractionEngine(prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"))
+csv_extraction_engine = CsvExtractionEngine(
+    prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"), column_names=["Name", "Age", "Address"]
+)
 
 # Define some unstructured data and columns
 csv_data = """
@@ -13,15 +15,12 @@
 Charlie is 40 and lives in Texas.
 """
 
-columns = ["Name", "Age", "Address"]
-
 
 # Create an agent and add the ExtractionTask to it
 agent = Agent()
 agent.add_task(
     ExtractionTask(
         extraction_engine=csv_extraction_engine,
-        args={"column_names": columns},
     )
 )
 
diff --git a/docs/griptape-framework/structures/src/tasks_7.py b/docs/griptape-framework/structures/src/tasks_7.py
index 909d00084..1a3a32b29 100644
--- a/docs/griptape-framework/structures/src/tasks_7.py
+++ b/docs/griptape-framework/structures/src/tasks_7.py
@@ -8,6 +8,7 @@
 # Instantiate the json extraction engine
 json_extraction_engine = JsonExtractionEngine(
     prompt_driver=OpenAiChatPromptDriver(model="gpt-3.5-turbo"),
+    template_schema=Schema({"users": [{"name": str, "age": int, "location": str}]}).json_schema("UserSchema"),
 )
 
 # Define some unstructured data and a schema
@@ -15,13 +16,12 @@
 Alice (Age 28) lives in New York.
 Bob (Age 35) lives in California.
 """
-user_schema = Schema({"users": [{"name": str, "age": int, "location": str}]}).json_schema("UserSchema")
+
 
 agent = Agent()
 agent.add_task(
     ExtractionTask(
         extraction_engine=json_extraction_engine,
-        args={"template_schema": user_schema},
     )
 )
 
diff --git a/griptape/artifacts/json_artifact.py b/griptape/artifacts/json_artifact.py
index b292879a9..26f69da7f 100644
--- a/griptape/artifacts/json_artifact.py
+++ b/griptape/artifacts/json_artifact.py
@@ -7,11 +7,11 @@
 
 from griptape.artifacts import BaseArtifact
 
-Json = Union[dict[str, "Json"], list["Json"], str, int, float, bool, None]
-
 
 @define
 class JsonArtifact(BaseArtifact):
+    Json = Union[dict[str, "Json"], list["Json"], str, int, float, bool, None]
+
     value: Json = field(converter=lambda v: json.loads(json.dumps(v)), metadata={"serializable": True})
 
     def to_text(self) -> str:
diff --git a/griptape/engines/extraction/csv_extraction_engine.py b/griptape/engines/extraction/csv_extraction_engine.py
index 40c9058ec..70977f1f8 100644
--- a/griptape/engines/extraction/csv_extraction_engine.py
+++ b/griptape/engines/extraction/csv_extraction_engine.py
@@ -17,7 +17,7 @@
 
 @define
 class CsvExtractionEngine(BaseExtractionEngine):
-    column_names: list[str] = field(default=Factory(list), kw_only=True)
+    column_names: list[str] = field(kw_only=True)
     system_template_generator: J2 = field(default=Factory(lambda: J2("engines/extraction/csv/system.j2")), kw_only=True)
     user_template_generator: J2 = field(default=Factory(lambda: J2("engines/extraction/csv/user.j2")), kw_only=True)
 
diff --git a/griptape/engines/extraction/json_extraction_engine.py b/griptape/engines/extraction/json_extraction_engine.py
index d3928c3ea..c817efd5f 100644
--- a/griptape/engines/extraction/json_extraction_engine.py
+++ b/griptape/engines/extraction/json_extraction_engine.py
@@ -18,9 +18,9 @@
 
 @define
 class JsonExtractionEngine(BaseExtractionEngine):
-    JSON_PATTERN = r"(?s)(\{.*\}|\[.*\])"
+    JSON_PATTERN = r"(?s)[^\[]*(\[.*\])"
 
-    template_schema: dict = field(default=Factory(dict), kw_only=True)
+    template_schema: dict = field(kw_only=True)
     system_template_generator: J2 = field(
         default=Factory(lambda: J2("engines/extraction/json/system.j2")), kw_only=True
     )
@@ -42,7 +42,7 @@ def json_to_text_artifacts(self, json_input: str) -> list[JsonArtifact]:
         json_matches = re.findall(self.JSON_PATTERN, json_input, re.DOTALL)
 
         if json_matches:
-            return [JsonArtifact(json.loads(e)) for e in json_matches]
+            return [JsonArtifact(e) for e in json.loads(json_matches[-1])]
         else:
             return []
 
diff --git a/griptape/templates/engines/extraction/json/system.j2 b/griptape/templates/engines/extraction/json/system.j2
index b6bac028a..987ff19a9 100644
--- a/griptape/templates/engines/extraction/json/system.j2
+++ b/griptape/templates/engines/extraction/json/system.j2
@@ -1,5 +1,3 @@
-Your answer MUST be JSON that successfully validates against the Extraction Template JSON Schema.
-
 Extraction Template JSON Schema: """{{ json_template_schema }}"""
 
 {% if rulesets %}
diff --git a/griptape/templates/engines/extraction/json/user.j2 b/griptape/templates/engines/extraction/json/user.j2
index 9b0a611f0..984977d9a 100644
--- a/griptape/templates/engines/extraction/json/user.j2
+++ b/griptape/templates/engines/extraction/json/user.j2
@@ -1,4 +1,4 @@
-Extract information from the Text.
+Extract information from the Text based on the Extraction Template JSON Schema into an array of JSON objects.
 Text: """{{ text }}"""
 
-JSON:
+JSON array:
diff --git a/tests/unit/engines/extraction/test_json_extraction_engine.py b/tests/unit/engines/extraction/test_json_extraction_engine.py
index 938438af4..b4aa58b75 100644
--- a/tests/unit/engines/extraction/test_json_extraction_engine.py
+++ b/tests/unit/engines/extraction/test_json_extraction_engine.py
@@ -21,15 +21,16 @@ def engine(self):
     def test_extract_text(self, engine):
         result = engine.extract_text("foo")
 
-        assert len(result.value) == 1
-        assert result.value[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
+        assert len(result.value) == 2
+        assert result.value[0].value == {"test_key_1": "test_value_1"}
+        assert result.value[1].value == {"test_key_2": "test_value_2"}
 
     def test_chunked_extract_text(self, engine):
         large_text = Path(normpath(join(dirname(__file__), "../../../resources", "test.txt"))).read_text()
 
         extracted = engine.extract_text(large_text * 50)
-        assert len(extracted) == 177
-        assert extracted[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
+        assert len(extracted) == 354
+        assert extracted[0].value == {"test_key_1": "test_value_1"}
 
     def test_extract_error(self, engine):
         engine.template_schema = lambda: "non serializable"
@@ -37,8 +38,10 @@ def test_extract_error(self, engine):
             engine.extract_text("foo")
 
     def test_json_to_text_artifacts(self, engine):
-        extracted = engine.json_to_text_artifacts('[{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]')
-        assert extracted[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
+        assert [
+            a.value
+            for a in engine.json_to_text_artifacts('[{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]')
+        ] == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
 
     def test_json_to_text_artifacts_no_matches(self, engine):
         assert engine.json_to_text_artifacts("asdfasdfasdf") == []
diff --git a/tests/unit/tools/test_extraction_tool.py b/tests/unit/tools/test_extraction_tool.py
index edf1663d2..1dbb9def2 100644
--- a/tests/unit/tools/test_extraction_tool.py
+++ b/tests/unit/tools/test_extraction_tool.py
@@ -39,14 +39,16 @@ def test_json_extract_artifacts(self, json_tool):
             {"values": {"data": {"memory_name": json_tool.input_memory[0].name, "artifact_namespace": "foo"}}}
         )
 
-        assert len(result.value) == 1
-        assert result.value[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
+        assert len(result.value) == 2
+        assert result.value[0].value == {"test_key_1": "test_value_1"}
+        assert result.value[1].value == {"test_key_2": "test_value_2"}
 
     def test_json_extract_content(self, json_tool):
         result = json_tool.extract({"values": {"data": "foo"}})
 
-        assert len(result.value) == 1
-        assert result.value[0].value == [{"test_key_1": "test_value_1"}, {"test_key_2": "test_value_2"}]
+        assert len(result.value) == 2
+        assert result.value[0].value == {"test_key_1": "test_value_1"}
+        assert result.value[1].value == {"test_key_2": "test_value_2"}
 
     def test_csv_extract_artifacts(self, csv_tool):
         csv_tool.input_memory[0].store_artifact("foo", TextArtifact("foo,bar\nbaz,maz"))

From dca34985688c819e60889da75bb85005bcc90629 Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Wed, 21 Aug 2024 09:23:51 -0700
Subject: [PATCH 05/10] Improve ListArtifact

---
 CHANGELOG.md                               |  2 ++
 griptape/artifacts/list_artifact.py        | 23 +++++++++++++---------
 griptape/schemas/base_schema.py            |  6 +++++-
 griptape/tasks/tool_task.py                |  6 +++++-
 tests/unit/artifacts/test_list_artifact.py |  6 ++++++
 5 files changed, 32 insertions(+), 11 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 59947dc5f..2288c9fc2 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,6 +7,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## Unreleased
 ### Added
 - `BaseConversationMemory.prompt_driver` for use with autopruning. 
+- Generic type support to `ListArtifact`.
+- Iteration support to `ListArtifact`.
 
 ### Fixed
 - Parsing streaming response with some OpenAi compatible services.
diff --git a/griptape/artifacts/list_artifact.py b/griptape/artifacts/list_artifact.py
index 298f29c6a..9ebec394c 100644
--- a/griptape/artifacts/list_artifact.py
+++ b/griptape/artifacts/list_artifact.py
@@ -1,23 +1,25 @@
 from __future__ import annotations
 
-from typing import TYPE_CHECKING, Optional
+from typing import TYPE_CHECKING, Generic, Optional, TypeVar
 
 from attrs import Attribute, define, field
 
 from griptape.artifacts import BaseArtifact
 
 if TYPE_CHECKING:
-    from collections.abc import Sequence
+    from collections.abc import Iterator, Sequence
+
+T = TypeVar("T", bound=BaseArtifact)
 
 
 @define
-class ListArtifact(BaseArtifact):
-    value: Sequence[BaseArtifact] = field(factory=list, metadata={"serializable": True})
+class ListArtifact(BaseArtifact, Generic[T]):
+    value: Sequence[T] = field(factory=list, metadata={"serializable": True})
     item_separator: str = field(default="\n\n", kw_only=True, metadata={"serializable": True})
     validate_uniform_types: bool = field(default=False, kw_only=True, metadata={"serializable": True})
 
     @value.validator  # pyright: ignore[reportAttributeAccessIssue]
-    def validate_value(self, _: Attribute, value: list[BaseArtifact]) -> None:
+    def validate_value(self, _: Attribute, value: list[T]) -> None:
         if self.validate_uniform_types and len(value) > 0:
             first_type = type(value[0])
 
@@ -31,18 +33,21 @@ def child_type(self) -> Optional[type]:
         else:
             return None
 
-    def __getitem__(self, key: int) -> BaseArtifact:
+    def __getitem__(self, key: int) -> T:
         return self.value[key]
 
     def __bool__(self) -> bool:
         return len(self) > 0
 
+    def __add__(self, other: BaseArtifact) -> ListArtifact[T]:
+        return ListArtifact(self.value + other.value)
+
+    def __iter__(self) -> Iterator[T]:
+        return iter(self.value)
+
     def to_text(self) -> str:
         return self.item_separator.join([v.to_text() for v in self.value])
 
-    def __add__(self, other: BaseArtifact) -> BaseArtifact:
-        return ListArtifact(self.value + other.value)
-
     def is_type(self, target_type: type) -> bool:
         if self.value:
             return isinstance(self.value[0], target_type)
diff --git a/griptape/schemas/base_schema.py b/griptape/schemas/base_schema.py
index f25e8870b..b285d1476 100644
--- a/griptape/schemas/base_schema.py
+++ b/griptape/schemas/base_schema.py
@@ -2,7 +2,7 @@
 
 from abc import ABC
 from collections.abc import Sequence
-from typing import Any, Literal, Union, _SpecialForm, get_args, get_origin
+from typing import Any, Literal, TypeVar, Union, _SpecialForm, get_args, get_origin
 
 import attrs
 from marshmallow import INCLUDE, Schema, fields
@@ -56,6 +56,10 @@ def _get_field_for_type(cls, field_type: type) -> fields.Field | fields.Nested:
 
         field_class, args, optional = cls._get_field_type_info(field_type)
 
+        # Resolve TypeVars to their bound type
+        if isinstance(field_class, TypeVar):
+            field_class = field_class.__bound__
+
         if attrs.has(field_class):
             if ABC in field_class.__bases__:
                 return fields.Nested(PolymorphicSchema(inner_class=field_class), allow_none=optional)
diff --git a/griptape/tasks/tool_task.py b/griptape/tasks/tool_task.py
index 6dd5000b3..68260ea91 100644
--- a/griptape/tasks/tool_task.py
+++ b/griptape/tasks/tool_task.py
@@ -84,7 +84,11 @@ def run(self) -> BaseArtifact:
             subtask.after_run()
 
             if isinstance(subtask.output, ListArtifact):
-                self.output = subtask.output[0]
+                first_artifact = subtask.output[0]
+                if isinstance(first_artifact, BaseArtifact):
+                    self.output = first_artifact
+                else:
+                    self.output = ErrorArtifact(f"Output is not an Artifact: {type(subtask.output[0])}")
             else:
                 self.output = InfoArtifact("No tool output")
         except Exception as e:
diff --git a/tests/unit/artifacts/test_list_artifact.py b/tests/unit/artifacts/test_list_artifact.py
index 06d234645..37769e1b8 100644
--- a/tests/unit/artifacts/test_list_artifact.py
+++ b/tests/unit/artifacts/test_list_artifact.py
@@ -23,6 +23,12 @@ def test___add__(self):
         assert artifact.value[0].value == "foo"
         assert artifact.value[1].value == "bar"
 
+    def test___iter__(self):
+        assert [a.value for a in ListArtifact([TextArtifact("foo"), TextArtifact("bar")])] == ["foo", "bar"]
+
+    def test_type_var(self):
+        assert ListArtifact[TextArtifact]([TextArtifact("foo")]).value[0].value == "foo"
+
     def test_validate_value(self):
         with pytest.raises(ValueError):
             ListArtifact([TextArtifact("foo"), BlobArtifact(b"bar")], validate_uniform_types=True)

From a7f10d4231ddc01769cc1b5128b632657087804d Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Thu, 22 Aug 2024 09:30:37 -0700
Subject: [PATCH 06/10] Make type covariant

---
 griptape/artifacts/list_artifact.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/griptape/artifacts/list_artifact.py b/griptape/artifacts/list_artifact.py
index 9ebec394c..02dd295cd 100644
--- a/griptape/artifacts/list_artifact.py
+++ b/griptape/artifacts/list_artifact.py
@@ -9,7 +9,7 @@
 if TYPE_CHECKING:
     from collections.abc import Iterator, Sequence
 
-T = TypeVar("T", bound=BaseArtifact)
+T = TypeVar("T", bound=BaseArtifact, covariant=True)
 
 
 @define

From b1ded7b2980c679f107a253b52ffd958c21fa906 Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Thu, 22 Aug 2024 09:46:55 -0700
Subject: [PATCH 07/10] Wording improvements

---
 CHANGELOG.md                                          | 2 +-
 docs/griptape-framework/engines/extraction-engines.md | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 498499d73..f78aa0751 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -11,7 +11,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Iteration support to `ListArtifact`.
 
 ### Changed
-- **BREAKING**: Split `BaseExtractionEngine.extract` into `extract` and `extract_artifacts` for consistency with `BaseSummaryEngine`.
+- **BREAKING**: Split `BaseExtractionEngine.extract` into `extract_text` and `extract_artifacts` for consistency with `BaseSummaryEngine`.
 - **BREAKING**: `BaseExtractionEngine` no longer catches exceptions and returns `ErrorArtifact`s.
 - **BREAKING**: `JsonExtractionEngine.template_schema` is now required.
 - **BREAKING**: `CsvExtractionEngine.column_names` is now required.
diff --git a/docs/griptape-framework/engines/extraction-engines.md b/docs/griptape-framework/engines/extraction-engines.md
index 43333e4b0..c9be21dde 100644
--- a/docs/griptape-framework/engines/extraction-engines.md
+++ b/docs/griptape-framework/engines/extraction-engines.md
@@ -10,7 +10,7 @@ As of now, Griptape supports two types of Extraction Engines: the CSV Extraction
 
 ## CSV
 
-The CSV Extraction Engine is designed for extracting CSV-formatted content from unstructured data.
+The CSV Extraction Engine extracts tabular content from unstructured data.
 
 ```python
 --8<-- "docs/griptape-framework/engines/src/extraction_engines_1.py"
@@ -24,7 +24,7 @@ Charlie,40,Texas
 
 ## JSON
 
-The JSON Extraction Engine is designed for extracting JSON-formatted content from unstructed data. 
+The JSON Extraction Engine extracts JSON-formatted content from unstructed data. 
 
 
 ```python

From e15ba3fae8d7cce7d5e517a865c3307d35011009 Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Thu, 22 Aug 2024 11:58:50 -0700
Subject: [PATCH 08/10] Better words

---
 docs/griptape-framework/engines/extraction-engines.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/griptape-framework/engines/extraction-engines.md b/docs/griptape-framework/engines/extraction-engines.md
index c9be21dde..cd06d22bc 100644
--- a/docs/griptape-framework/engines/extraction-engines.md
+++ b/docs/griptape-framework/engines/extraction-engines.md
@@ -10,7 +10,7 @@ As of now, Griptape supports two types of Extraction Engines: the CSV Extraction
 
 ## CSV
 
-The CSV Extraction Engine extracts tabular content from unstructured data.
+The CSV Extraction Engine extracts tabular content from unstructured text.
 
 ```python
 --8<-- "docs/griptape-framework/engines/src/extraction_engines_1.py"
@@ -24,7 +24,7 @@ Charlie,40,Texas
 
 ## JSON
 
-The JSON Extraction Engine extracts JSON-formatted content from unstructed data. 
+The JSON Extraction Engine extracts JSON-formatted content from unstructed text. 
 
 
 ```python

From f66925ca3e8308bd369530927e73c9419666f20f Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Thu, 3 Oct 2024 11:37:47 -0700
Subject: [PATCH 09/10] Fix typo

---
 docs/griptape-framework/engines/extraction-engines.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/griptape-framework/engines/extraction-engines.md b/docs/griptape-framework/engines/extraction-engines.md
index cd06d22bc..c00352691 100644
--- a/docs/griptape-framework/engines/extraction-engines.md
+++ b/docs/griptape-framework/engines/extraction-engines.md
@@ -24,7 +24,7 @@ Charlie,40,Texas
 
 ## JSON
 
-The JSON Extraction Engine extracts JSON-formatted content from unstructed text. 
+The JSON Extraction Engine extracts JSON-formatted content from unstructured text. 
 
 
 ```python

From d1250094caf57d439909d137303428f0d23ab9f8 Mon Sep 17 00:00:00 2001
From: Collin Dutter <collin@griptape.ai>
Date: Thu, 3 Oct 2024 11:39:44 -0700
Subject: [PATCH 10/10] Fix bad merge

---
 griptape/artifacts/list_artifact.py | 12 ------------
 1 file changed, 12 deletions(-)

diff --git a/griptape/artifacts/list_artifact.py b/griptape/artifacts/list_artifact.py
index 16fc323df..02dd295cd 100644
--- a/griptape/artifacts/list_artifact.py
+++ b/griptape/artifacts/list_artifact.py
@@ -18,18 +18,6 @@ class ListArtifact(BaseArtifact, Generic[T]):
     item_separator: str = field(default="\n\n", kw_only=True, metadata={"serializable": True})
     validate_uniform_types: bool = field(default=False, kw_only=True, metadata={"serializable": True})
 
-    def __getitem__(self, key: int) -> T:
-        return self.value[key]
-
-    def __bool__(self) -> bool:
-        return len(self) > 0
-
-    def __add__(self, other: BaseArtifact) -> ListArtifact[T]:
-        return ListArtifact(self.value + other.value)
-
-    def __iter__(self) -> Iterator[T]:
-        return iter(self.value)
-
     @value.validator  # pyright: ignore[reportAttributeAccessIssue]
     def validate_value(self, _: Attribute, value: list[T]) -> None:
         if self.validate_uniform_types and len(value) > 0: