Models

ToolResult

Standardized result for all action executions. Backed by a Rust struct via PyO3.

Fields

Field	Type	Default	Description
success	bool	True	Whether the execution was successful
message	str	""	Human-readable result description
prompt	Optional[str]	None	Suggestion for AI about next steps
error	Optional[str]	None	Error message when success is False
context	Dict[str, Any]	{}	Additional context data

Methods

Method	Returns	Description
with_error(error)	ToolResult	Create copy with error info (sets success=False)
with_context(**kwargs)	ToolResult	Create copy with updated context
to_dict()	Dict[str, Any]	Convert to dictionary
to_json()	str	Serialize to a JSON string
__eq__(other)	bool	Equality comparison
__str__()	str	Human-readable string
__repr__()	str	Unambiguous representation

json.dumps() is not supported directly

ToolResult is a Rust-backed object and cannot be passed to json.dumps() directly. Use to_json() or convert to a dict first:

import json
result = success_result("done")

# Option 1 — built-in JSON serializer (recommended, uses Rust serde)
json_str = result.to_json()

# Option 2 — via dict
json_str = json.dumps(result.to_dict())

# Option 3 — serialize_result (supports JSON and MsgPack)
from dcc_mcp_core import serialize_result
json_str = serialize_result(result)

Factory Functions

from dcc_mcp_core import success_result, error_result, from_exception, validate_action_result, ToolResult

# Success result with context
result = success_result("Created 5 spheres", prompt="Use modify_spheres next", count=5)

# Error result with possible solutions
error = error_result(
    "Failed", "File not found",
    prompt="Check path",
    possible_solutions=["Verify file exists", "Check permissions"],
    path="/bad/path",
)

# From exception string
exc_result = from_exception(
    "ValueError: bad input",
    message="Import failed",
    include_traceback=True,
)

# Validate/normalize any value to ToolResult
validate_action_result(result)                          # pass-through
validate_action_result({"success": True, "message": "OK"})  # dict → ARM
validate_action_result("hello")                         # wrap as success

Factory Function Signatures

Function	Signature	Description
success_result	(message, prompt=None, **context) -> ToolResult	Create a successful result
error_result	(message, error, prompt=None, possible_solutions=None, **context) -> ToolResult	Create a failed result
from_exception	(error_message, message=None, prompt=None, include_traceback=True, possible_solutions=None, **context) -> ToolResult	Wrap an exception as a result
validate_action_result	(result: Any) -> ToolResult	Normalize dict/str/None/ToolResult → ToolResult

SkillMetadata

Metadata parsed from SKILL.md frontmatter. All fields are readable and writable.

Fields

Field	Type	Default	Description
name	str	—	Unique identifier
description	str	""	Human-readable description
tools	list[ToolDeclaration]	[]	Declared tools parsed from SKILL.md
dcc	str	"python"	Target DCC application
tags	list[str]	[]	Classification tags
scripts	list[str]	[]	Discovered script file paths
skill_path	str	""	Absolute path to skill directory
version	str	"1.0.0"	Skill version
depends	list[str]	[]	Names of required skills
metadata_files	list[str]	[]	Files in metadata/ directory
license	str	""	License identifier
compatibility	str	""	Compatibility string
allowed_tools	list[str]	[]	Allowed tool names (restricts which tools can be registered)
groups	list[SkillGroup]	[]	Tool groups for progressive exposure