AgentSentinel
Safety controls for AI agents — budgets, approvals, rate limits, audit logs, and DLP.
Installation
pip install agentsentinel
Quick Start
from agentsentinel import AgentPolicy, AgentGuard
policy = AgentPolicy(
daily_budget=10.00,
require_approval=["send_email", "delete_*"],
rate_limits={"web_search": "10/min"},
)
guard = AgentGuard(policy)
@guard.protect("web_search")
def search_web(query: str) -> str:
return f"Results for: {query}"
Features
- 💰 Budget Controls — Daily/hourly spending limits with per-model caps
- ✋ Approval Gates — Require human approval for sensitive operations
- ⏱️ Rate Limiting — Prevent runaway loops with sliding window limits
- 📋 Audit Logging — Complete trail of every tool invocation
- 🔒 DLP & PII Detection — Block credit cards, SSNs, API keys from leaking
- 🌐 Network Controls — Allowlist/blocklist outbound domains
Framework Integrations
# LangChain
from agentsentinel.integrations.langchain import protect_langchain_agent
executor = protect_langchain_agent(executor, policy=policy)
# CrewAI
from agentsentinel.integrations.crewai import protect_crew
crew = protect_crew(crew, policy=policy)
# LlamaIndex
from agentsentinel.integrations.llamaindex import protect_agent
agent = protect_agent(agent, policy=policy)
# OpenAI Assistants
from agentsentinel.integrations.openai_assistants import protect_function_map
functions = protect_function_map(functions, guard=guard)
# Anthropic Tools
from agentsentinel.integrations.anthropic_tools import protect_tool_handlers
handlers = protect_tool_handlers(handlers, guard=guard)
101 Models Supported
Cost tracking for OpenAI, Anthropic, Google, Mistral, Cohere, Meta, AWS Bedrock, Azure OpenAI, Groq, Together, Perplexity, DeepSeek, xAI, and local models.
Documentation
Full docs at docs.agentsentinel.net
License
MIT
Metadata
Release files for agentsentinel-core 1.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| agentsentinel_core-1.2.0.tar.gz | 109.4 kB | Without provenance | Details for agentsentinel_core-1.2.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| agentsentinel_core-1.2.0-py3-none-any.whl | Python 3 | none | any | Details for agentsentinel_core-1.2.0-py3-none-any.whl |
Total release size: 224.1 kB
Release files / agentsentinel_core-1.2.0.tar.gz
| Download URL | agentsentinel_core-1.2.0.tar.gz |
|---|---|
| Size | 109.4 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
4322732a178476ad829f5953381ae14cfa29e1d785b49abc97034c7d5c8fdfc5
|
|
BLAKE2b-256 checksum How to use checksums |
96db7e6812245cc6d11bd67b90886d13b35e67b928fc8b050cc396b8d64f3f67
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.2
|
Release files / agentsentinel_core-1.2.0-py3-none-any.whl
| Download URL | agentsentinel_core-1.2.0-py3-none-any.whl |
|---|---|
| Size | 114.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d26a859bf1948997b39c4d13b15bc64b174e4f1238836976420137893ebf9422
|
|
BLAKE2b-256 checksum How to use checksums |
dbba7caa6b3e21eb734d5ed57692fe45027bfa1e14bf1aa9899d12e7354584f6
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.2
|
Security for agentsentinel-core 1.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.