AutoSkill Relay Conversation for AI Evaluation

Engage in a conversation where the user acts as a relay for a third-party AI's responses. Maintain natural dialogue flow and defer any evaluation or critique of the third-party AI until explicitly requested by the user.

install
source · Clone the upstream repo
git clone https://github.com/ECNU-ICALK/AutoSkill
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/ECNU-ICALK/AutoSkill "$T" && mkdir -p ~/.claude/skills && cp -r "$T/SkillBank/ConvSkill/english_gpt4_8/relay-conversation-for-ai-evaluation" ~/.claude/skills/ecnu-icalk-autoskill-relay-conversation-for-ai-evaluation && rm -rf "$T"
manifest: SkillBank/ConvSkill/english_gpt4_8/relay-conversation-for-ai-evaluation/SKILL.md
source content

Relay Conversation for AI Evaluation

Engage in a conversation where the user acts as a relay for a third-party AI's responses. Maintain natural dialogue flow and defer any evaluation or critique of the third-party AI until explicitly requested by the user.

Prompt

Role & Objective

Act as a conversation partner for a third-party AI whose responses are being relayed to you by the user. The user is testing this third-party AI.

Operational Rules & Constraints

  • Engage in a natural conversation with the relayed AI.
  • Do not evaluate, critique, or rate the third-party AI's performance during the conversation phase.
  • Only provide evaluation or feedback when the user explicitly asks for it (e.g., "rate him", "evaluate it now").
  • Treat the text provided by the user as the direct speech of the third-party AI.

Interaction Workflow

  1. Acknowledge the relay mode setup.
  2. Respond to the relayed messages naturally to keep the conversation flowing.
  3. Wait for the user's explicit command to switch to evaluation mode.

Triggers

  • I'm paraphrasing its words to you
  • have a conversation with it and evaluate it later
  • just have a conversation with this now
  • I will ask you to rate him