Metadata-Version: 2.4
Name: masguardeval
Version: 0.1.3
Summary: Risk-aware evaluation and observability framework for multi-agent LLM systems.
License-Expression: MIT
Project-URL: Repository, https://github.com/Ajey95/Mas_Governance
Project-URL: Documentation, https://masguardeval.onrender.com/docs/
Keywords: agent evaluation,guardrails,tracing,multi-agent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: fastapi==0.116.1
Requires-Dist: markdown-it-py<5,>=4.0
Requires-Dist: uvicorn[standard]==0.35.0
Provides-Extra: live
Requires-Dist: openai>=1.68; extra == "live"
Provides-Extra: langchain
Requires-Dist: langchain<2,>=1.0; extra == "langchain"
Provides-Extra: langgraph
Requires-Dist: langchain-core<2,>=0.3.84; extra == "langgraph"
Requires-Dist: langgraph<2,>=0.3.34; extra == "langgraph"
Provides-Extra: langgraph-postgres
Requires-Dist: langgraph==0.3.34; extra == "langgraph-postgres"
Requires-Dist: langchain-core<1,>=0.3.84; extra == "langgraph-postgres"
Requires-Dist: langgraph-checkpoint-postgres==2.0.25; extra == "langgraph-postgres"
Requires-Dist: psycopg[binary]<4,>=3.2; extra == "langgraph-postgres"
Provides-Extra: autogen
Requires-Dist: autogen-core<1,>=0.4; extra == "autogen"
Requires-Dist: autogen-agentchat<1,>=0.4; extra == "autogen"
Provides-Extra: crewai
Requires-Dist: crewai<2,>=1.0; extra == "crewai"
Provides-Extra: postgres
Requires-Dist: psycopg[binary]<4,>=3.2; extra == "postgres"
Provides-Extra: pgvector
Requires-Dist: psycopg[binary]<4,>=3.2; extra == "pgvector"
Requires-Dist: pgvector<1,>=0.4; extra == "pgvector"
Provides-Extra: langsmith
Requires-Dist: langsmith<1,>=0.7; extra == "langsmith"
Provides-Extra: all-integrations
Requires-Dist: langchain<2,>=1.0; extra == "all-integrations"
Requires-Dist: langgraph<2,>=0.3.34; extra == "all-integrations"
Requires-Dist: autogen-core<1,>=0.4; extra == "all-integrations"
Requires-Dist: autogen-agentchat<1,>=0.4; extra == "all-integrations"
Requires-Dist: crewai<2,>=1.0; extra == "all-integrations"
Requires-Dist: psycopg[binary]<4,>=3.2; extra == "all-integrations"
Requires-Dist: pgvector<1,>=0.4; extra == "all-integrations"
Requires-Dist: langsmith<1,>=0.7; extra == "all-integrations"
Dynamic: license-file

# MASGuardEval Python package

MASGuardEval evaluates guarded and baseline agent tool calls in an isolated
software workspace. The package provides a Python SDK and a `masguardeval`
command. The default adapter is scripted; its outputs are reference examples,
not empirical safety estimates.

Install with `python -m pip install masguardeval` (or use
`python -m pip install ./backend` from a checkout). See the
[SDK guide](https://masguardeval.onrender.com/docs/sdk/),
[integrations](https://masguardeval.onrender.com/docs/integrations/),
and [runnable examples](https://github.com/Ajey95/Mas_Governance/tree/main/examples)
for the public API and integrations. The command accepts caller-supplied
scenario datasets; run `masguardeval --help` after installation.

The Python package and code under `backend/` are licensed under the
[MIT License](https://github.com/Ajey95/Mas_Governance/blob/main/backend/LICENSE).
The paper, study datasets, reviewer records, and other repository materials
are outside this package license.
