Register agent types and enforce result schemas in the Sub-agent Dispatch harness.
npx skills add https://github.com/AesopScott/mojo --skill agent-types-aesopscott
Please help me install this Agent Skill. Skill: agent-types Source: https://github.com/AesopScott/mojo/tree/main/harnesses/skills/agent-types Command: npx skills add https://github.com/AesopScott/mojo --skill agent-types-aesopscott
This skill addresses the complexity of managing sub-agent behaviors, preventing capability creep, and ensuring that parallel execution workflows remain bounded and predictable.
Use the agent-types skill to inspect the current sub-agent dispatch harness and propose a new capability configuration for the upcoming sprint.
High-intent search queries and answers about installing and using this skill.
Preventing capability creep in multi-agent systems requires registering agent types with defined authority boundaries. You can configure pluggable planning strategies to ensure parallel execution workflows remain bounded, predictable, and restricted to specific tools.
Enforcing result schemas for parallel workers standardizes output shapes to ensure seamless merging. By defining strict validation rules for both positive and boundary test cases within the dispatch harness, you maintain consistent operational boundaries across agent workflows.
Sub-agent dispatch is the mechanism of routing tasks to specialized worker types while enforcing authority limits. You need it for complex agent workflows when managing parallel execution, standardizing results, and preventing unauthorized access to sensitive system functions.
Yes, you can configure pluggable planning strategies for specific agent types. This allows you to define custom parallel execution parameters and authority boundaries, ensuring each registered worker operates strictly within its designated operational limits.
Restricting worker access to sensitive functions involves defining authority limits during agent type registration. By enforcing result schemas and validating boundary test cases, the dispatch harness ensures workers only interact with their explicitly permitted tools.
The primary limitation of a sub-agent dispatch harness is the strict adherence required to defined authority limits and validation rules. Without properly registering boundary test cases and result schemas, parallel execution workflows may fail to merge outputs seamlessly.
AesopScottConfigure risk-tiered proof requirements and waiver shortcuts in the Proof Authority harness.
carson-sweetExecute standardized test scenarios to validate SweetClaude agent behavioral contracts.
Kabilan108Manage concurrent processes across tmux sessions and panes.
PatkikImplement template method and pluggable strategies for Laravel workflows.
