Delegation-Chain Tracking
When an agent invokes another agent, maintain an auditable chain-of-delegation that preserves the original principal and intermediate authority.
This page is the complete instruction page for one pattern called "Delegation-Chain Tracking." It explains the problem this pattern solves, the idea behind the solution, when you should (and should not) use it, and what happens afterward, both the good effects and the costs.
This matters because building AI agents is not just about making them clever. It is also about making them safe and predictable. Following a well-tested pattern like this one helps avoid common mistakes, and shows you exactly which safety rules and regulations it connects to, listed under "Standards Mesh" on this page.
Context
Multi-agent workflows can obscure "on whose behalf" an action is taken. A specialist agent acting on a supervisor's delegation, itself acting for a user, produces three layers of authority. Without tracking, the audit trail collapses.
Problem
Agent-to-agent delegation without chain-tracking loses the principal of record.
Forces
- Clean agent boundaries vs end-to-end accountability
- Chain depth vs metadata overhead
Solution
Propagate a delegation chain header with every inter-agent call: [original principal → supervising agent → worker agent → ... ]. Each hop appends its identity and timestamp. Authorisation decisions consider the full chain, not just the immediate caller. Observability reconstructs the chain end-to-end.
Where this pattern acts in the two lifecycles
05 Observe. Tool result ingestion, state update. 7 of 61 catalogue patterns act at this stage.
05 Operate & Monitor. Runtime telemetry, drift detection. 18 of 61 catalogue patterns act at this stage.
Stages and counts from the catalogue data. The Runtime lens is the loop the agent runs each turn; the AgentOps lens is the lifecycle the team runs the agent through. See the Matrix for every pattern at each stage.
Applicability
- Hierarchical multi-agent systems
- Agent-as-Tool compositions
- Cross-organisation agent invocation (A2A/MCP)
Anti-Patterns
- Each agent hop re-authenticating as itself, losing principal context
- Chains that grow unbounded without audit
Consequences
- +End-to-end accountability
- +Supports principal-aware authorisation
- −Header-propagation discipline required
- −Adds audit-record volume
