mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-22 16:25:58 +00:00
b78ff50d8d
Port from Kilo-Org/kilocode#11955: Gemini validates every object schema's required list strictly against the same node's properties and fails the ENTIRE GenerateContentRequest with HTTP 400 'required[0]: property is not defined' when a name has no matching property. MCP servers (e.g. the GitHub remote MCP) routinely emit array item schemas carrying required without properties, which made every request on the native Gemini path fail before any model output. sanitize_gemini_schema() now filters required to names present in the node's properties and drops the keyword when nothing valid remains. Applies recursively (properties / items / anyOf). Tool handlers still validate required fields at execution time, so nothing the model could actually use is lost. Scoped to the Gemini-facing sanitizer only — the universal tools/schema_sanitizer.py already prunes typed object nodes, and its remaining gap (untyped nodes) is contested by open PR #20151.
227 lines
9 KiB
Python
227 lines
9 KiB
Python
"""Tests for agent.gemini_schema — OpenAI→Gemini tool parameter translation."""
|
|
|
|
from agent.gemini_schema import (
|
|
sanitize_gemini_schema,
|
|
sanitize_gemini_tool_parameters,
|
|
)
|
|
|
|
|
|
class TestSanitizeGeminiSchema:
|
|
def test_strips_unknown_top_level_keys(self):
|
|
"""$schema / additionalProperties etc. must not reach Gemini."""
|
|
schema = {
|
|
"type": "object",
|
|
"$schema": "https://json-schema.org/draft/2020-12/schema",
|
|
"additionalProperties": False,
|
|
"properties": {"foo": {"type": "string"}},
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert "$schema" not in cleaned
|
|
assert "additionalProperties" not in cleaned
|
|
assert cleaned["type"] == "object"
|
|
assert cleaned["properties"] == {"foo": {"type": "string"}}
|
|
|
|
def test_preserves_string_enums(self):
|
|
"""String-valued enums are valid for Gemini and must pass through."""
|
|
schema = {"type": "string", "enum": ["pending", "done", "cancelled"]}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["type"] == "string"
|
|
assert cleaned["enum"] == ["pending", "done", "cancelled"]
|
|
|
|
def test_drops_integer_enum_to_satisfy_gemini(self):
|
|
"""Gemini rejects int-typed enums; the sanitizer must drop the enum.
|
|
|
|
Regression for the Discord tool's ``auto_archive_duration``:
|
|
``{type: integer, enum: [60, 1440, 4320, 10080]}`` caused
|
|
Gemini HTTP 400 INVALID_ARGUMENT
|
|
"Invalid value ... (TYPE_STRING), 60" on every request that
|
|
shipped the full tool catalog to generativelanguage.googleapis.com.
|
|
"""
|
|
schema = {
|
|
"type": "integer",
|
|
"enum": [60, 1440, 4320, 10080],
|
|
"description": "Minutes (60, 1440, 4320, 10080).",
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["type"] == "integer"
|
|
assert "enum" not in cleaned
|
|
# description must survive so the model still sees the allowed values
|
|
assert cleaned["description"].startswith("Minutes")
|
|
|
|
def test_drops_number_enum(self):
|
|
"""Same rule applies to ``type: number``."""
|
|
schema = {"type": "number", "enum": [0.5, 1.0, 2.0]}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["type"] == "number"
|
|
assert "enum" not in cleaned
|
|
|
|
def test_drops_boolean_enum(self):
|
|
"""And to ``type: boolean`` (Gemini rejects non-string entries)."""
|
|
schema = {"type": "boolean", "enum": [True, False]}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["type"] == "boolean"
|
|
assert "enum" not in cleaned
|
|
|
|
def test_keeps_string_enum_even_when_numeric_values_coexist_as_strings(self):
|
|
"""Stringified-numeric enums ARE valid for Gemini; don't drop them."""
|
|
schema = {"type": "string", "enum": ["60", "1440", "4320", "10080"]}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["enum"] == ["60", "1440", "4320", "10080"]
|
|
|
|
def test_drops_nested_integer_enum_inside_properties(self):
|
|
"""The fix must apply recursively — the Discord case is nested."""
|
|
schema = {
|
|
"type": "object",
|
|
"properties": {
|
|
"auto_archive_duration": {
|
|
"type": "integer",
|
|
"enum": [60, 1440, 4320, 10080],
|
|
"description": "Thread archive duration in minutes.",
|
|
},
|
|
"status": {
|
|
"type": "string",
|
|
"enum": ["active", "archived"],
|
|
},
|
|
},
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
props = cleaned["properties"]
|
|
# Integer enum is dropped...
|
|
assert props["auto_archive_duration"]["type"] == "integer"
|
|
assert "enum" not in props["auto_archive_duration"]
|
|
# ...but the sibling string enum is preserved.
|
|
assert props["status"]["enum"] == ["active", "archived"]
|
|
|
|
def test_drops_integer_enum_inside_array_items(self):
|
|
"""Array item schemas recurse through ``items``."""
|
|
schema = {
|
|
"type": "array",
|
|
"items": {"type": "integer", "enum": [1, 2, 3]},
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["items"]["type"] == "integer"
|
|
assert "enum" not in cleaned["items"]
|
|
|
|
def test_non_dict_input_returns_empty(self):
|
|
assert sanitize_gemini_schema(None) == {}
|
|
assert sanitize_gemini_schema("not a schema") == {}
|
|
assert sanitize_gemini_schema([1, 2, 3]) == {}
|
|
|
|
|
|
class TestRequiredPropertyPruning:
|
|
"""Gemini rejects ``required`` names missing from the node's ``properties``.
|
|
|
|
Regression for the Kilo-Org/kilocode#11955 bug class: MCP servers (e.g.
|
|
the GitHub remote MCP) emit array item schemas whose ``required`` lists
|
|
reference properties that don't exist in the same node — Google fails the
|
|
entire GenerateContentRequest with HTTP 400 "property is not defined".
|
|
"""
|
|
|
|
def test_drops_required_when_node_has_no_properties(self):
|
|
schema = {"type": "object", "required": ["a", "b"]}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert "required" not in cleaned
|
|
|
|
def test_filters_ghost_required_entries(self):
|
|
schema = {
|
|
"type": "object",
|
|
"properties": {"x": {"type": "string"}},
|
|
"required": ["x", "ghost"],
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["required"] == ["x"]
|
|
|
|
def test_prunes_inside_array_items(self):
|
|
"""The exact shape from the GitHub MCP report — nested in items."""
|
|
schema = {
|
|
"type": "object",
|
|
"properties": {
|
|
"issue_fields": {
|
|
"type": "array",
|
|
"items": {
|
|
"type": "object",
|
|
"required": ["field_id", "value"],
|
|
},
|
|
},
|
|
},
|
|
"required": ["issue_fields"],
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
items = cleaned["properties"]["issue_fields"]["items"]
|
|
assert "required" not in items
|
|
# Top-level required is valid and survives.
|
|
assert cleaned["required"] == ["issue_fields"]
|
|
|
|
def test_prunes_node_without_explicit_type(self):
|
|
"""Nodes carrying properties+required but no ``type`` key still prune."""
|
|
schema = {
|
|
"properties": {"x": {"type": "string"}},
|
|
"required": ["x", "ghost"],
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["required"] == ["x"]
|
|
|
|
def test_valid_required_untouched(self):
|
|
schema = {
|
|
"type": "object",
|
|
"properties": {"a": {"type": "string"}, "b": {"type": "integer"}},
|
|
"required": ["a", "b"],
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["required"] == ["a", "b"]
|
|
|
|
def test_drops_non_string_required_entries(self):
|
|
schema = {
|
|
"type": "object",
|
|
"properties": {"a": {"type": "string"}},
|
|
"required": ["a", 42, None],
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["required"] == ["a"]
|
|
|
|
def test_prunes_inside_anyof_branches(self):
|
|
schema = {
|
|
"anyOf": [
|
|
{
|
|
"type": "object",
|
|
"properties": {"x": {"type": "string"}},
|
|
"required": ["x", "ghost"],
|
|
},
|
|
{"type": "object", "required": ["orphan"]},
|
|
]
|
|
}
|
|
cleaned = sanitize_gemini_schema(schema)
|
|
assert cleaned["anyOf"][0]["required"] == ["x"]
|
|
assert "required" not in cleaned["anyOf"][1]
|
|
|
|
|
|
class TestSanitizeGeminiToolParameters:
|
|
def test_empty_parameters_return_valid_object_schema(self):
|
|
"""Gemini requires ``parameters`` to be a valid object schema."""
|
|
cleaned = sanitize_gemini_tool_parameters({})
|
|
assert cleaned == {"type": "object", "properties": {}}
|
|
|
|
def test_discord_create_thread_parameters_no_longer_trip_gemini(self):
|
|
"""End-to-end regression: the exact shape that was rejected in prod."""
|
|
params = {
|
|
"type": "object",
|
|
"properties": {
|
|
"action": {"type": "string", "enum": ["create_thread"]},
|
|
"auto_archive_duration": {
|
|
"type": "integer",
|
|
"enum": [60, 1440, 4320, 10080],
|
|
"description": "Thread archive duration in minutes "
|
|
"(create_thread, default 1440).",
|
|
},
|
|
},
|
|
"required": ["action"],
|
|
}
|
|
cleaned = sanitize_gemini_tool_parameters(params)
|
|
aad = cleaned["properties"]["auto_archive_duration"]
|
|
# The field that triggered the Gemini 400 is gone.
|
|
assert "enum" not in aad
|
|
# Type + description survive so the model still knows what to send.
|
|
assert aad["type"] == "integer"
|
|
assert "1440" in aad["description"]
|
|
# And the string-enum sibling is untouched.
|
|
assert cleaned["properties"]["action"]["enum"] == ["create_thread"]
|