Prepend a non-negotiable constitution to every sub-agent prompt.
npx skills add https://github.com/vertexcover-io/tarash --skill constitution-vertexcover-io
Please help me install this Agent Skill. Skill: constitution Source: https://github.com/vertexcover-io/tarash/tree/main/.claude/skills/constitution Command: npx skills add https://github.com/vertexcover-io/tarash --skill constitution-vertexcover-io
The Constitution provides non-negotiable guardrails that are automatically applied to every sub-agent context, ensuring consistency, scope discipline, and safe coding practices across the orchestration layer.
Load the constitution into the orchestrator so every sub-agent prompt includes the safety preamble.
High-intent search queries and answers about installing and using this skill.
To enforce safety guardrails on sub-agent prompts, you can prepend a non-negotiable constitution to every sub-agent context, ensuring consistent scope discipline and safe coding practices across the orchestration layer.
Prompt injection prevention for autonomous agents involves applying deterministic constitution preambles to sub-agent prompts, constraining scope and requiring explicit plan-based approvals to reduce risk in autonomous decision making.
You apply a constitution to all sub-agent prompts by loading it into the orchestrator, which automatically prepends the safety preamble to guarantee compliance with defined governance rules during execution.
Yes, you can enforce explicit approvals for plan-based changes by applying a constitution that enforces scope boundaries and requires approvals, enabling fast reporting when rules are violated within orchestrated workflows.
Autonomous agents ignore scope boundaries when lacking enforced governance; applying a non-negotiable constitution prepend to sub-agent prompts constrains scope and ensures deterministic, auditable behavior.
Yes, the constitution approach works without external dependencies, applying guardrails directly within the orchestration layer to maintain governance and reduce risk in autonomous multi-agent systems.
abzhawDesign and optimize LLM prompts for Julia's orchestrator and sub-agents.
JCE-Joshhh77Grade, merge, and validate sub-agent outputs before acting.
ek33450505Create and validate structured feature plans with Agent Dispatch Manifest (ADM) blocks for /orchestrate execution.
Kgan01Enforce constitutional safety rules across autonomous agent workflows.
