oopsyz

Use this skill when working with the bundled single-file `scripts/codex_ws_client.py` WebSocket client for `codex app-server`, including sending prompts, REPL use, resumed threads, JSON output, approval handling, tracing, or debugging Codex app-server protocol behavior.

Use this skill when working with the bundled `scripts/claude_cli_client.py` wrapper for the local `claude` CLI, including one-shot prompts, resumed sessions, REPL use, structured JSON output, or trace/debugging of Claude stream-json output.

Use this skill when the user wants an intelligent multi-turn discussion or brainstorm on a topic. Claude actively participates — sharing what it agrees with, what it disagrees with, and introducing out-of-the-box angles — while the peer model is driven either through the Codex WebSocket client or the Claude brainstorm adapter. The result is a real back-and-forth between two AI perspectives, not just a relay.

Use this skill when the user wants an intelligent multi-turn discussion or brainstorm on a topic. Claude actively participates — sharing what it agrees with, what it disagrees with, and introducing out-of-the-box angles — while Codex responds via the codex-ws-client WebSocket transport. The result is a real back-and-forth between two AI perspectives, not just a relay.

Use this skill when the user wants to manage, validate, or navigate the ea/sa/da/dev enterprise convention artifacts. Triggers include: adding an initiative, domain, workstream, implementation, or roadmap item; validating referential integrity; tracing a routing chain; or reviewing the status of any layer. Keywords: initiative, domain, workstream, implementation, roadmap, ea, sa, da, domain-registry, domain-workstreams, domain-implementations, domain-roadmap, convention, routing, referential integrity.

Analyze a recruiter email or LinkedIn message to detect scams, phishing attempts, and fraudulent job opportunities. Use this skill whenever the user pastes a recruiter email, LinkedIn message, or job offer and wants to know if it's legitimate. Trigger on phrases like "is this legit", "check this recruiter", "verify this job offer", "is this a scam", "LinkedIn message", "recruiter email", or any time the user shares unsolicited job communication for review.