Skip to content

Aegis

The simplest way to govern AI agent actions. No infra. No lock-in. Just Python.

pip install agent-aegis → YAML policy → governance in 5 minutes. Works with LangChain, CrewAI, OpenAI, Anthropic, MCP, and more.

Try it live in your browser — no install needed.

The Problem

AI agents are getting real-world access — calling APIs, modifying databases, running shell commands, browsing the web. Without governance, a single hallucination can:

  • Bulk-delete your CRM contacts
  • Submit wrong forms to government portals
  • Trigger irreversible API calls at 3am
  • Run up cloud bills with infinite loops

There's no sudo for AI agents. Until now.

What is Aegis?

Aegis is a Python library (not a platform, not a server) that wraps every AI agent action with policy checks, approval gates, and audit logging:

Action      Policy        Approval       Execute     Audit
  |            |              |              |           |
read CRM  --> auto (low)  --> skip -------> run ------> logged
bulk edit --> approve (high) --> human y/n -> run ------> logged
delete *  --> block (critical) ------------> X --------> logged

Copy, paste, run — zero config needed:

from aegis import Action, Policy

policy = Policy.from_dict({
    "version": "1",
    "defaults": {"risk_level": "low", "approval": "auto"},
    "rules": [{"name": "block_delete", "match": {"type": "delete_*"},
               "risk_level": "critical", "approval": "block"}]
})

safe = policy.evaluate(Action(type="read_users", target="db"))
print(safe.approval)   # Approval.AUTO  ✅

danger = policy.evaluate(Action(type="delete_users", target="db"))
print(danger.approval)  # Approval.BLOCK 🚫

Or with a YAML file — 3 lines:

from aegis import Action, Policy, Runtime

runtime = Runtime(executor=your_executor, policy=Policy.from_yaml("policy.yaml"))
result = await runtime.run_one(Action("delete", "crm", params={"id": "all"}))
# Policy checked. Approval gated. Audit logged. Done.

Key Features

Feature Description
YAML policies Glob matching, first-match-wins, JSON Schema for validation
Smart conditions time_after, time_before, weekdays, param_gt/lt/eq/contains/matches
4-tier risk model low / medium / high / critical with per-rule overrides
7 approval handlers CLI, Slack, Discord, Telegram, email, webhook, or custom
Audit trail SQLite + JSONL + webhook + Python logging — compliance-ready
7 framework adapters LangChain, CrewAI, OpenAI Agents SDK, Anthropic Claude, Playwright, httpx, MCP
REST API server aegis serve policy.yaml — govern from any language
CLI tools aegis init, validate, simulate, audit, stats, serve
Type-safe Full mypy --strict, py.typed marker
< 1ms overhead Policy evaluation adds minimal latency to your agent

Real-World Use Cases

Scenario How Aegis Helps
CRM Agent Read contacts freely, review updates, block mass deletes
Code Agent Read files safely, review edits, block deploys to production
Financial Agent View accounts, approve payments over threshold, block after-hours transfers
Browser Agent Navigate freely, review form fills, block JavaScript execution
Data Pipeline SELECT freely, review INSERTs to production, block DROP TABLE
Compliance Every action logged with full context for SOC2/GDPR/HIPAA evidence

Install

pip install agent-aegis
aegis init  # Generate a starter policy

With integrations:

pip install 'agent-aegis[langchain]'      # LangChain
pip install 'agent-aegis[crewai]'         # CrewAI
pip install 'agent-aegis[openai-agents]'  # OpenAI Agents SDK
pip install 'agent-aegis[anthropic]'      # Anthropic Claude
pip install 'agent-aegis[httpx]'          # REST APIs
pip install 'agent-aegis[playwright]'     # Browser automation
pip install 'agent-aegis[server]'         # REST API server
pip install 'agent-aegis[all]'            # Everything

Quick Example

from aegis import Action, Policy, Runtime
from aegis.adapters.httpx_adapter import HttpxExecutor

async with Runtime(
    executor=HttpxExecutor(base_url="https://api.example.com"),
    policy=Policy.from_yaml("policy.yaml"),
) as runtime:
    # Single action
    result = await runtime.run_one(Action("get", "/users"))

    # Multiple actions with plan
    plan = runtime.plan([
        Action("get", "/users"),              # auto-execute (low risk)
        Action("post", "/users", params={"json": {"name": "Alice"}}),  # approve
        Action("delete", "/users/all"),       # blocked (critical risk)
    ])
    results = await runtime.execute(plan)

Production Ready

Aspect Detail
518 tests, 92% coverage Every adapter, handler, and edge case tested
Type-safe mypy --strict with zero errors
< 1ms evaluation Policy check adds negligible overhead
Fail-safe Blocked actions never execute, period
Audit immutability Results are frozen; audit writes happen before returning
Zero external deps Core has no required infrastructure

Roadmap

Version Status Features
0.1 Released Policy engine, 7 adapters (incl. MCP), CLI, audit, conditions, JSON Schema
0.1.3 Released REST API server, retry/rollback, dry-run, hot-reload, policy merge, 7 approval handlers, runtime hooks
0.1.4 Released Multi-agent foundations, PolicyHierarchy, performance optimizations, security hardening
0.2 Q2 2026 Dashboard UI, rate limiting, queue-based async execution
0.3 Q3 2026 Agent identity, cross-agent audit correlation
0.4 Q4 2026 Multi-agent governance, centralized policy server
1.0 2027 Distributed governance, policy versioning, multi-tenant API