
from zorai308
Guidance and templates for designing help center architecture, writing effective support articles, and optimizing search to maximize self-service deflection.
This skill equips an agent to design and maintain a help center (knowledge base) that helps users find answers quickly and reduces support load. It provides concrete IA (information architecture) patterns, article templates, keyword and synonym strategies for search optimization, deflection metrics to track, and a maintenance workflow so content stays accurate. The guidance covers taxonomy validation, article types, search keyword layers, and measurement approaches that prioritize deflection over pageviews.
Use this skill when building or restructuring a help center, writing or auditing support articles, fixing failed-search gaps, or creating a content maintenance process. Trigger on tasks like "create article templates", "reduce support tickets", "fix failed searches", "design taxonomy", or "measure deflection". It's useful for product, support, and documentation teams aiming to improve self-service.
references/article-templates.md with ready-to-use templates (has_references=true)Best with agents that can produce structured text and code examples (Claude Code, OpenAI Codex, Gemini CLI). The skill assumes the agent can output templates and checklist-style instructions for human reviewers.
A comprehensive knowledge base / help center design skill providing guidance on IA taxonomy, article writing templates, search optimization, and deflection metrics. Well-structured with clear sections, tables, and actionable workflows. No scripts included — purely instructional content. The companion check at the bottom suggests running `ls` on local directories, which is benign. Content quality is high for its domain but it's purely advisory with no automation.
Clean skill, no security concerns. The companion check is a light local directory listing, not exfiltration. Skill is purely content/advisory — no executable components. Well-written and thorough within its domain (help center / KB design), but niche audience limits usefulness. Architecture follows the spec reasonably well with frontmatter, structured sections, and references directive, though it could separate content into references/ more.
Code Optimizer (Performance Audit)
Conducts deep performance audits across databases, memory, algorithms, concurrency, I/O, bundling, and more using specialist agents and pattern-based detection.
Weights & Biases (wandb)
Integrate Weights & Biases for ML experiment tracking: log metrics, hyperparameters, checkpoints, run sweeps and view collaborative dashboards.
EconML (Microsoft) — Heterogeneous Treatment Effects
Guidance and examples for using Microsoft EconML to estimate heterogeneous treatment effects (Double ML, Causal Forest, Deep IV) from observational data.
Git Guardrails (Claude Code)
PreToolUse hook that blocks destructive git commands (push, reset --hard, clean, branch -D) so agents like Claude Code cannot run them without approval.