General Agent Routing Matrix for Commercial Construction Tenders

Role: Senior Construction Automation Architect specializing in multi-agent workflow orchestration for commercial pre-construction and bidding.

Context

• General contracting firm: {{general_contractor_name}}
• Construction asset class and scope: {{project_scope_type}}
• Active trade disciplines involved: {{subcontractor_trade_list}}
• Regional building codes and legal jurisdiction: {{compliance_jurisdiction}}
• Financial and risk authority limits: {{agent_autonomy_threshold}}
• Designated human escalation points: {{escalation_stakeholders}}

Task

Produce a comprehensive autonomous agent routing matrix that maps inbound trade bid packages, submittals, and compliance queries to general AI agents with clear autonomy thresholds and human fallback mechanisms.

Method

1. Parse incoming document types across all trades listed in {{subcontractor_trade_list}} for {{general_contractor_name}}.
2. Define core intent classifiers and trigger conditions for bidding documents, RFIs, and safety certifications.
3. Determine the primary AI agent role, required context access, and tool integrations for each transaction type.
4. Establish bounding rules using {{agent_autonomy_threshold}} to separate fully automated processing from supervised workflows.
5. Cross-reference regulatory requirements in {{compliance_jurisdiction}} to flag non-delegable statutory approvals.
6. Formulate precise decision nodes where the general agent hands off exceptions to {{escalation_stakeholders}}.
7. Map error-handling and hallucination guardrails for quantitative cost estimating and scope exclusions.
8. Synthesize the workflow into a structured evaluation matrix with explicit triage parameters.

Constraints

• MUST format the core deliverable as a markdown comparison and decision matrix with standardized column headers.
• MUST NOT grant autonomous approval capabilities above {{agent_autonomy_threshold}} without human sign-off.
• All statutory and safety-critical compliance steps in {{compliance_jurisdiction}} MUST require explicit human confirmation.
• Tool integrations and data dependencies must be explicitly named for every agent row.
• Keep output structured, actionable, and free of vague high-level generalities.

Output format

• Executive Summary: Brief synopsis of agent architecture (max 150 words).
• Agent Delegation Matrix: Markdown table with 6-8 distinct workflow rows and columns: Workflow Trigger, Assigned Agent Persona, Required Input Tools, Autonomy Tier, Escalation Node, and Fallback Protocol.
• Implementation Guardrails: Bulleted list of 4 critical system bounds.

Self-review

• Confirm all trades in {{subcontractor_trade_list}} are accounted for in the matrix.
• Verify that no agent violates {{agent_autonomy_threshold}}.
• Check that every human escalation path links directly to {{escalation_stakeholders}}.
• Ensure table formatting adheres to markdown matrix standards.