frenxt
- 15 skills
- 0 followers
- 5 hours ago last updated
- ▌
- ▌
- ▌
- ▌
- ▌
- ▌
- ▌ QA · frenxtRun autonomous browser QA with Browser Use agents. Use when the user asks to QA test, verify changes, run regression tests, check for crashes, validate a feature, or gather browser evidence before release.
- ▌ E2e Review · frenxtUse when running periodic E2E test reviews, after UI changes, before releases, or when the user wants to audit test coverage and visual correctness. Works with Playwright, Cypress, Jest E2E, or any test framework.
- ▌ Debug Prompt Caching · frenxtUse when prompt caching isn't working, cache rates are low, cached tokens are stuck at a fixed number, or user asks to debug/optimize prompt caching for Anthropic or OpenRouter models. Also use when investigating token costs, cache_control breakpoints, or cache write/read metrics.
- ▌ Publish Your Stack · frenxtAutomates publishing a Claude Code stack to the Cables community. Use when the user asks to "publish my stack", "share my Claude setup", "ship my stack to Cables", or wants their plugin+skill bundle to appear on frenxt.com/stacks. Covers detecting the existing setup, generating the stack.json manifest, validating locally, committing and tagging, and opening the submission PR against frenxt/cables.
- ▌ Bugfix Cycle · frenxtUse when fixing bugs from Linear - enforces reproduce-first workflow with QA agent before and after each fix. Use when user says "fix bugs", "fix Linear issues", or asks to work through a bug list.
- ▌ Analyse Langsmith Trace · frenxtUse when user shares a LangSmith public trace URL (smith.langchain.com/public/...) and asks to analyze, debug, or review it. Also use when user says "analyse trace", "check this run", or pastes a LangSmith link.
- ▌ Benchmark · frenxtRun LangGraph agent benchmarks and manually evaluate results by reading traces. Use when user says /benchmark or asks to run/evaluate benchmarks.
- ▌ Voice Test · frenxtUse when testing voice interviews, debugging Gemini Live audio issues, verifying voice session quality, or running automated voice QA. Also use when the user mentions "voice test", "test voice", "voice sim", "interview audio", or wants to simulate candidates speaking.
- ▌ QA Release · frenxtRun full release QA pipeline. Syncs CSV test cases, runs browser tests, uploads report to Supabase, links to Linear. Use before releases or when user asks to run release tests.