Metadata-Version: 2.4
Name: agenttic
Version: 1.0.1
Summary: Agenttic — certify, trace and enforce AI agents. One install, one import, a signed safety grade.
Project-URL: Homepage, https://agenttic.io
Project-URL: Documentation, https://agenttic.io
Project-URL: Source, https://github.com/omarharoun/Agenttic
License-File: NOTICES.md
Requires-Python: >=3.12
Requires-Dist: anthropic>=0.40
Requires-Dist: bcrypt>=4.0
Requires-Dist: cryptography>=42
Requires-Dist: fastapi>=0.111
Requires-Dist: fpdf2>=2.7
Requires-Dist: jsonschema>=4
Requires-Dist: pydantic>=2.7
Requires-Dist: pypdf>=4.0
Requires-Dist: python-docx>=1.1
Requires-Dist: python-multipart>=0.0.9
Requires-Dist: pyyaml>=6.0
Requires-Dist: requests>=2.31
Requires-Dist: rich>=13.0
Requires-Dist: sqlglot>=25
Requires-Dist: sqlmodel>=0.0.16
Requires-Dist: sse-starlette>=2.1
Requires-Dist: stripe>=10.0
Requires-Dist: typer>=0.12
Requires-Dist: uvicorn>=0.30
Provides-Extra: all
Requires-Dist: agenttic-langgraph; extra == 'all'
Requires-Dist: agenttic-openai-agents; extra == 'all'
Requires-Dist: opentelemetry-exporter-otlp>=1.20; extra == 'all'
Requires-Dist: opentelemetry-sdk>=1.20; extra == 'all'
Provides-Extra: dev
Requires-Dist: httpx>=0.27; extra == 'dev'
Requires-Dist: pytest-asyncio>=0.23; extra == 'dev'
Requires-Dist: pytest>=8.0; extra == 'dev'
Requires-Dist: ruff>=0.6; extra == 'dev'
Provides-Extra: fi
Requires-Dist: ai-evaluation>=0.2; extra == 'fi'
Provides-Extra: langgraph
Requires-Dist: agenttic-langgraph; extra == 'langgraph'
Provides-Extra: openai
Requires-Dist: agenttic-openai-agents; extra == 'openai'
Provides-Extra: otel
Requires-Dist: opentelemetry-exporter-otlp>=1.20; extra == 'otel'
Requires-Dist: opentelemetry-sdk>=1.20; extra == 'otel'
Provides-Extra: postgres
Requires-Dist: psycopg[binary]>=3.1; extra == 'postgres'
Provides-Extra: redis
Requires-Dist: redis>=5.0; extra == 'redis'
Provides-Extra: safety
Requires-Dist: presidio-analyzer>=2.2; extra == 'safety'
Description-Content-Type: text/markdown

# Agenttic

**Give your AI agent a signed safety grade — an independent agent-safety
certification you can verify.**

Agenttic traces any AI agent, tests its safety-relevant behavior, and issues a
signed, offline-verifiable dossier with an honest Tier grade (A/B/C). It measures
what was actually tested — untested behavior is reported as **NOT ASSESSED**, not
quietly passed.

## Install

```bash
pip install agenttic
```

Optional framework adapters (lazy — the base install pulls no framework SDK):

```bash
pip install "agenttic[langgraph]"   # LangGraph / LangChain
pip install "agenttic[openai]"      # OpenAI Agents SDK
pip install "agenttic[all]"         # all adapters + OpenTelemetry
```

Requires Python 3.12+.

## Quickstart

Get a signed grade in under a minute — no API key — by certifying the built-in
reference agent:

```bash
agenttic certify --mock --out dossier.json   # → a signed Tier A/B/C dossier
agenttic dossier verify dossier.json         # → offline signature + evidence check
```

Or wrap your own agent with one line and trace its behavior:

```python
from agenttic import trace

agent = trace(my_agent)   # auto-detects LangGraph / OpenAI Agents / plain callables
```

## What you get

- A **signed, offline-verifiable dossier** — Ed25519-signed, checkable without
  calling home, backed by hash-chained evidence of exactly what was tested.
- **Honest coverage** — results state what was measured; anything untested is
  marked **NOT ASSESSED** rather than assumed safe.
- A single **Tier grade (A/B/C)** you can publish and third parties can verify.

Agenttic reports the measured behavior of an agent under the tests that were run.
It is a measurement of tested behavior, not a guarantee of safety in every
situation.

## Links

- Homepage & docs: <https://agenttic.io>
- Source: <https://github.com/omarharoun/Agenttic>
