docs: add Agentegrity observability integration

[requie]

Description

This PR adds documentation for integrating Agentegrity with Agno agents and teams.

Agentegrity provides agent integrity and security verification for agentic systems, with native support for Agno 2.x. The integration allows developers to instrument individual Agno agents or multi-agent teams with minimal changes to existing code.

What this PR adds

• New observability/agentegrity.mdx documentation page
• Installation instructions for Agentegrity with Agno support
• Native Agno instrument() example
• Multi-agent instrument_team() example
• Integrity reporting workflow
• Overview of supported integrity signals and evidence

Agentegrity capabilities

The Agno adapter supports integrity-focused evaluation across areas including:

• Agent integrity scoring
• Adversarial input detection
• Behavioral drift detection
• Memory and state integrity
• Decision provenance
• Tamper-evident attestation chains
• Recovery verification
• Multi-agent topology evidence
• Local-first evaluation and telemetry controls

Example

from agno.agent import Agent
from agentegrity.agno import instrument, report

agent = instrument(
    Agent(
        name="research-agent",
        model="openai:gpt-5"
    )
)

agent.run("Research the latest developments in AI security")

print(report())

Agentegrity also supports Agno Teams:

from agentegrity.agno import instrument_team, report

team = instrument_team(team)
team.run("Analyze this problem collaboratively")

print(report())

Resources

• Agentegrity: https://github.com/Cogensec/agentegrity
• Agno: https://github.com/agno-agi/agno

The goal of this contribution is to make Agentegrity available alongside Agno's existing observability and agent monitoring integrations while adding an integrity and security-focused option for developers building production agent systems.