Skip to main content

The trust layer for AI agents. Observability, identity, and reputation in one line of code.

Project description

authe

The trust layer for AI agents. Observability, identity, and reputation in one line of code.

PyPI License: MIT

Install

pip install authe

Quick Start

import authe
authe.init()

# That's it. Your agent is now observable.
# View actions at dashboard.authe.me

What it does

authe.init() automatically instruments your agent and captures:

  • Every tool call — function name, inputs, outputs, duration
  • LLM requests — model, message count, tool usage
  • File operations — writes tracked with path and mode
  • System commands — subprocess calls captured
  • Scope violations — alerts when agent exceeds declared capabilities

All actions are sent to your dashboard at dashboard.authe.me with a tamper-proof audit trail.

Configuration

import authe

authe.init(
    api_key="ak_xxxxx",          # or set AUTHE_API_KEY env var
    agent_name="my-agent",        # auto-detected if not set
    capabilities=["read:email", "write:file"],  # declared permissions
    redact_pii=True,              # redact sensitive fields
)

Environment Variables

Variable Description
AUTHE_API_KEY Your API key (required)
AUTHE_AGENT_NAME Agent name (optional, auto-detected)

Manual Tracking

For custom tools that aren't auto-detected:

from authe.instrumentor import track

@track("send_email")
def send_email(to, subject, body):
    # your code here
    pass

Or track directly:

from authe import get_client

client = get_client()
client.track_action(
    tool="custom_tool",
    input_data={"key": "value"},
    output_data={"result": "done"},
    status="success",
)

Supported Frameworks

Framework Auto-instrumented
OpenAI
LangChain
CrewAI 🔜
AutoGPT 🔜
Custom agents via @track decorator

How it works

  1. authe.init() registers your agent and gets a short-lived token
  2. Actions are captured via monkey-patching (zero code changes needed)
  3. Actions are batched and sent to the API every 5 seconds
  4. Your dashboard shows a real-time timeline of everything your agent did
  5. Scope violations trigger alerts automatically

Links

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

authe-0.1.0.tar.gz (9.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

authe-0.1.0-py3-none-any.whl (10.4 kB view details)

Uploaded Python 3

File details

Details for the file authe-0.1.0.tar.gz.

File metadata

  • Download URL: authe-0.1.0.tar.gz
  • Upload date:
  • Size: 9.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.7

File hashes

Hashes for authe-0.1.0.tar.gz
Algorithm Hash digest
SHA256 33b2b1464cf2076c233fe82e6f43ca6b5d5deab5daaf1cec8da53db5ed06a549
MD5 64101482570d6f0497157ec98b4f5f98
BLAKE2b-256 17a8376b252e9394f4ebec8f0194c9eb19a1e8c450131c90cfc6b0c181acb68f

See more details on using hashes here.

File details

Details for the file authe-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: authe-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 10.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.7

File hashes

Hashes for authe-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 37a24f6fbfb44646d77d11fc196ce6cc8c7c10792d1b61acf366135d320c64e0
MD5 dfee8c1f8e0b31f27db3dffc74c5dbd1
BLAKE2b-256 69cfb8a2498e3f22f3992bde679f57d71bfdaeb89e4570abe10ce715a3e6f7cf

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page