How does pre-response validation work?

The skill runs a validation checklist on every output before it is finalized, checking for banned voices, incorrect tool usage, and violations of style rules.

Does this work with other agents like Cursor or Codex?

Yes, through its integration with Ruler, the constraints defined in this skill are propagated to other agent configurations including Cursor, Codex, and Amp.

Can I add custom rules to the validator?

Yes, you can update the Frustration Index with new patterns, evidence, and persistence scores to tailor the validation logic to your specific workflow.

What is the Frustration Index?

The Frustration Index is a ranked list of AI behavioral patterns that historically fail to persist across sessions. This skill monitors these patterns to ensure your preferences stick forever.

Frustration Eradication

Name: Frustration Eradication
Author: plurigrid

byplurigrid

•

生产力与工作流

Enforces persistent user preferences and prevents repetitive AI behavioral regressions through a pre-response validation framework.

The Frustration Eradication skill acts as a persistent memory and validation layer for Claude Code, ensuring that specific user preferences remain enforced across all sessions. By implementing a 'Frustration Index,' it automatically checks every AI response against a set of ranked constraints—such as specific voice bans, tool mandates like Exa for semantic search, and strict verbosity limits. This skill effectively eliminates 'context drift' by validating responses before they are output, saving developers from the repetitive task of correcting the AI's behavior.

主要功能

01Strict control over response verbosity, preambles, and code comments

02Automatic propagation of rules across multiple agent configurations via Ruler

03GF(3) conservation checks for mathematical and compositional coherence

042 GitHub stars

05Pre-response validation checklist to catch behavioral regressions

06Enforcement of tool-specific mandates like Exa over generic web search

使用场景

01Ensuring the AI never uses banned voices or prohibited phrases in voice-enabled workflows

02Mandating parallel subagent execution for complex tasks to maximize performance

03Maintaining strict code style preferences like 'no comments' or 'no as-any' without manual reminders

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add plurigrid/asi frustration-eradication

For use in Claude.ai and ChatGPT

Download Skill