prompt-evaluation-runn…
Use when evaluating prompts, LLM outputs, red-team suites, or model behavior with local eval configs and safe provider/cost controls.
Use when a persistent local Gemini ACP session should keep context across multiple advisory turns.
$ npx -y skills add yeaight7/agent-powerups --skill relay-gemini --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/relay-geminiContext preview
The summary Claude sees to decide when to auto-load this skill.
Use when a persistent local Gemini ACP session should keep context across multiple advisory turns.
name: relay-gemini description: Use when a persistent local Gemini ACP session should keep context across multiple advisory turns.
Use Gemini through Agent Powerups relay for multi-turn advisory work where context continuity matters more than one-shot speed.
Compatible with: `claude-code`, `codex`, `generic`
Requires:
| Situation | Use | | --- | --- | | Multi-turn review or advisory dialogue | `relay` - context persists across turns | | Long-running code review with follow-ups | `relay` - same session, same context | | Single focused question, no follow-up needed | `ask-gemini` - simpler, no daemon | | Quick brainstorm, one answer expected | `ask-gemini` - lighter weight | | Cross-session context | `relay init` + `relay ask` with context file |
Do not use relay when a single `ask-gemini` turn would be enough.
1. Verify Gemini is available:
gemini --version
2. Create a context file when the session needs durable background:
apx relay init <name>
3. Start the session:
apx relay start <name> --provider gemini
4. Ask follow-up questions in the same session:
apx relay ask <name> "<prompt>"
5. Check session health when a turn fails or times out:
apx relay status <name>
6. Stop the session when the advisory loop is complete:
apx relay stop <name>
Curated power-ups for coding agents: skills, slash commands, MCP configs, hooks, AGENTS.md templates, and workflows for serious software engineering. Claude Code, Codex, Antigravity CLI, Cursor and more
Repo: yeaight7/agent-powerups
Use when evaluating prompts, LLM outputs, red-team suites, or model behavior with local eval configs and safe provider/cost controls.
Use when creating or reviewing red-team eval plugins, attack templates, grader rubrics, safety fixtures, or model-risk test metadata.
Use when designing, running, debugging, or hardening deterministic eval suites for agent skills, prompts, tool workflows, or MCP-backed cases.
Use when designing tool definitions for a new agent or subagent, an agent shows high retry rates, ambiguous tool invocations, or silent failures, or an…
Use when routing a prompt to a local provider CLI for a second opinion, review, or plan -- you are about to call a provider directly, need the response saved…
Use when starting work in an unfamiliar area of a codebase, spawning a subagent that needs targeted file context, a first search pass missed the relevant file,…