← ClaudeAtlas

tejlisted

Router for the Tej Stack skill suite. Use when the user invokes Tej Stack without naming a specific skill, or asks which Tej Stack skill fits a task. Triggers on tej, tej stack, tejstack, use tej, run tej stack on this, and speech-to-text aliases like tay stack, taj stack, tedge stack, tage stack, or tej stack. Routes feature audits to walkie-talkie, failure explanations to trace-failure, duplication sweeps to abstraction-police, performance hunts to big-o-police, AI answer-engine visibility audits to get-cited (and requests to fix SEO and AEO end to end to tej-core spotlight), slow LLM responses and time to first token to tej-core lag-doctor, requests to map or diagram a flow, architecture, or concept to x-ray, API bills, token spend, surprise cloud bills, and infrastructure cost to cost-police, and app promo videos, ads, and store preview videos to promo-reel, then hands off to that skill rather than doing the work itself.
varuntej07/tej-stack · ★ 1 · DevOps & Infrastructure · score 74
Install: claude install-skill varuntej07/tej-stack
# Tej > **Update check (run silently first):** run `bash "<this skill's directory>/../../bin/check-update.sh"` and read its single `TEJ_STACK:` status line. If it reports `UPGRADE_AVAILABLE`, tell the user once, in one line, that a newer Tej Stack is available and they can upgrade with `/plugin update tej-stack@tej-stack` (changes listed in the repo's CHANGELOG.md), then continue routing normally. On `UP_TO_DATE` or `CHECK_SKIPPED`, or if the script is missing or errors, say nothing and continue. The check must never block, delay, or fail the task. Send the request to the right Tej Stack skill and hand off. Do not perform the underlying audit here; this skill exists so a vague or voice-dictated invocation still lands on the right specialist. ## Routing table | The user's request is about | Route to | |---|---| | Whether a finished feature actually works for every user, actor, and recovery path | `walkie-talkie` | | A failure, error, crash, broken behavior, or "what happened and why" | `trace-failure` | | Duplicated code, drifted copies, repetitive-feeling modules, whether to merge or inline an abstraction | `abstraction-police` | | Slowness, memory, scale degradation, data-structure or algorithm choices | `big-o-police` | | Showing up in ChatGPT, Perplexity, Claude, Copilot, or Google AI answers; AEO, GEO, AI SEO, llms.txt, schema for AI | `get-cited` | | Slow LLM responses: chat or voice takes too long to answer, time to first token, p95 or p99 of a model call | `/tej-co