Browse reusable Agent Skills, each with a clear purpose and practical guidance.
Debug a user-reported issue, Linear ticket, or incident report by combining Datadog (APM, logs, metrics) with the Langfuse repo to establish a root cause. Use when given a Linear issue URL/ID (e.g. LFE-XXXX), a GitHub issue, or a pasted error/report and asked to investigate, root-cause, or triage. Produces a structured analysis — error breakdown, hypothesis-by-class, suggested patches with code references.
Seed local Langfuse test data with one command: large/branching observation trees (v3 and v4 events), long sessions, bulk traces for list performance. Use whenever a task needs ClickHouse/Postgres test data — e.g. "seed a complex trace", "make a tough session", "fill the trace list", "test v4 events UI", or when debugging trace/session/list rendering or performance. Never write ad-hoc seed scripts or raw ClickHouse inserts.
Decide whether open Bug PRs target the correct branch. A bug must be fixed on the lowest maintained branch where it exists, then merged up. Use when the user says "triage bug PRs", "which PRs need retargeting", or "retarget triage".
Review a change (a PR, the current branch diff, or a set of files) or audit a component or the whole tree for missing or incorrect security hardening. Reasons about trust boundaries from first principles, then checks the code against Symfony's hardening-invariant families and runs the .github/sa-tools gates. Use when the user says "security review", "security audit", "check hardening", "review this PR/branch for security", "audit <component> for <vuln class>", "is any hardening missing", or names a vulnerability class to hunt for.
Write and maintain Playwright end-to-end tests for the Onyx application. Use when creating new E2E tests, debugging test failures, adding test coverage, or when the user mentions Playwright, E2E tests, or browser testing.
Benchmark Dragonfly (and compare against Valkey/Redis) on local or remote cloud instances, then produce a performance + memory report with charts and a raw-data appendix. Use this whenever the user runs /benchmark, or asks to benchmark / load-test / measure throughput, QPS, latency, or memory efficiency of Dragonfly or another Redis-compatible server, to compare Dragonfly vs Valkey or Redis, or to generate a benchmark report. Trigger it even when the user just describes the goal ("see how fast Dragonfly is on this box", "compare memory per key vs Valkey") without saying the word "benchmark".
Reproduce AFL++ fuzz crashes from GitHub Actions. Use when user provides a GitHub Actions fuzz run URL and wants to reproduce and analyze the crash locally.
Interacts with live Hunk diff review sessions via CLI. Inspects review focus, navigates files and hunks, reloads session contents, and adds inline review comments. Use when the user has a Hunk session running or wants to review diffs interactively.
Instructions for how to validate changes made to Lighthouse. Must invoke when making changes to Lighthouse.
Generate an e2e profiling trace of an SGLang server run. Launches a server, validates accuracy, captures a Chrome-compatible trace, and returns the profile path.