Gate chat-agent prompts by scoring SMART POLE atoms before XML handoff.
npx skills add https://github.com/cuongpt083/smart-pole-skill --skill sp-chat-agent
Please help me install this Agent Skill. Skill: sp-chat-agent Source: https://github.com/cuongpt083/smart-pole-skill/tree/main/skills/sp-chat-agent Command: npx skills add https://github.com/cuongpt083/smart-pole-skill --skill sp-chat-agent
The SMART POLE Chat Enforcer gates vague chat-agent prompts, ensuring prompts meet a surgical readiness standard before releasing control and producing a machine-readable <master_prompt> XML handoff for downstream agents.
Provide a user prompt and allow the Enforcer to ask clarifying questions to extract SP-atoms, compute a readiness score, and emit the <master_prompt> block when ready.
High-intent search queries and answers about installing and using this skill.
To enforce prompt readiness, you need a gatekeeping mechanism that scores prompt completeness and withholds machine-readable handoffs until a threshold is met. This Skill analyzes prompts into atoms and gates the final XML handoff.
A SMART POLE atom is a structural unit extracted from a user prompt during analysis. The Enforcer isolates these atoms to score readiness, detect conflicts, and ensure locale handling before generating a structured XML handoff.
Generating a structured XML handoff requires scoring prompt readiness through clarifying interviews. Once the readiness threshold is achieved, the system emits a machine-readable <master_prompt> XML block for deterministic downstream parsing.
No, this prompt gating mechanism is not intended for coding-agent execution or direct code generation tasks. It specifically focuses on interactive prompt refinement and handoff orchestration for chat-agent workflows.
Yes, XML output gating integrates with system prompts and reference documents to support domain-aware context. This ensures policy-compliant executions and safe handoffs when the readiness threshold is reached.
Incomplete master-prompt handoffs occur when prompts lack surgical readiness. The Enforcer prevents premature handoffs by interviewing users, extracting SP-atoms, and withholding the final XML block until readiness scoring passes.
MihaiOnSoftwareLoad user rules from ~/.ai/rules/ in defined order.
treevesSurvey local AEM Edge Delivery Services blocks and Block Collection to build a consolidated inventory.
navapbcAutomate creation and refinement of domain-specific guidance for policy rule extraction.
faisalanjumTest direct invocation of agent-specific MCP tools to verify permissions.
