Agent Autonomy Tiers, Multi-Key Pooling, and Thought Signatures

1. Overview & Business Value#

SmartERP formalizes AI agent behavior using a rigorous five-tier operational autonomy matrix, ensuring complete clarity regarding which decisions can execute automatically versus those requiring explicit human sign-off.

1. Five-Tier Autonomy Classification Matrix#

2. Multi-Key Pooling and 429 Failover Mechanics#

High-throughput agents utilize a dynamic four-key round-robin pooler implemented in addons/muk_ai/providers/google.py. When an upstream API endpoint returns an HTTP 429 Rate Limit error, the pooler isolates the key for 60 seconds and immediately retries using the next healthy key in the sequence:

$$\text{ActiveKey} = \text{Keys}[(\text{Counter} \bmod N)]$$

Request ──> [Key Selector] ──> Key A (200 OK) ──> Return Response
                              Key B (429 Rate Limit) ──> Cooldown 60s
                                  └──> Failover ──> Key C (200 OK)

3. Cryptographic Thought Signatures#

For complex multi-turn reasoning and tool invocation, Gemini 3.x and Claude agents generate deterministic cryptographic thought signatures (thoughtSignature). These signatures prove that tool invocations were produced directly by the model's internal reasoning chain rather than synthetic parameter injection.

2. Prerequisites & Security Access Groups#

• Required Modules: muk_ai, muk_mcp, smarterp_ai_suite
• Security Group: Ashewa AI / User (Technical ID: muk_ai.group_ai_user)
• Master Data Dependencies: Configured AI provider credentials, active system prompts, and departmental cost centers.

3. Step-by-Step UI Execution & Verification#

Step 1: Navigating to the Operational Interface#

1. Log into your SmartERP instance at https://smarterp.et/odoo.
2. Open the main App Switcher and select Ashewa AI.
3. Select the operational workspace corresponding to this agent suite.

{screenshot:20-concepts-step1-navigation}

Step 2: Executing Agent Actions#

1. Review agent prompts and execution parameters.
2. Trigger the scheduled or on-demand workflow action.
3. Validate output in the generated operational log.

{screenshot:20-concepts-step2-execution}

4. Advanced Configuration Points#

• Scheduled Crons: muk_ai.ir_cron_agent_health_check runs every 30 minutes.
• Multi-Company Isolation: Multi-company record rules ensure agent context is isolated to Company 1 (Ashewa Technology Solutions S.C.).

5. Comprehensive Troubleshooting Matrix#