メニュー
Use this skill when communicating on Discord to decide whether to use emoji reactions or messages. Also use when asked about "reaction protocol", "discord efficiency", or how agents should communicate.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
This skill should be used when the user asks to "do a retro", "review team activity", "analyze delivery process", "identify issues from today", "daily retrospective", or when a reminder triggers for the daily retro. Analyzes Discord message history to identify delivery process issues and propose solutions.
Use this skill when asked to "create a new agent", "deploy a new agent", "add an agent", or set up a new team member. Also when someone asks "how do I create an agent" or needs to configure agent triggers, channels, or escalation.
Use this skill when asked to "generate an image", "create a picture", "make an illustration", "draw something", or when visual content needs to be created. Provides image generation via the gen-image CLI tool.
Use this skill when you need to restart an agent, increase cost limits, check agent status, or perform any agent maintenance tasks. Invoke with "/agent-maintenance" or when asked about restarting agents.
This skill should be used when the user asks to "test in browser", "open a webpage", "take a screenshot", "fill a form", "E2E test", "verify in browser", "check the UI", "test the frontend", or needs to interact with a web application. Provides browser automation capabilities via the `agent-browser` CLI.
| name | discord-reactions |
| description | Use this skill when communicating on Discord to decide whether to use emoji reactions or messages. Also use when asked about "reaction protocol", "discord efficiency", or how agents should communicate. |
Reduce unnecessary token usage by using emoji reactions for simple acknowledgments and status updates instead of full text messages.
Golden Rule: If it would take more than 5 words to explain, use a message. Otherwise, use a reaction.
| Emoji | Meaning | When to Use |
|---|---|---|
| 👀 | "Seen/Reading" | You've seen the message and are reviewing it |
| 👍 | "Acknowledged/OK" | Simple agreement or acknowledgment |
| ✅ | "Done/Complete" | Task completed successfully |
| ⏳ | "In Progress" | Working on it now |
| 🔄 | "Processing" | Thinking/investigating, will respond soon |
| 🚫 | "Can't do/Blocked" | Unable to proceed (follow up with why) |
| Emoji | Meaning | When to Use |
|---|---|---|
| 🎯 | "Claimed/Taking this" | You'll handle this task |
| 📋 | "Ticket created" | GitHub issue created |
| 🔍 | "Reviewing" | Code review or investigation in progress |
| 🧪 | "Testing" | Running tests or QA |
| 🚀 | "Deployed/Shipped" | Code merged or deployed |
| 📦 | "Ready for review" | Work complete, needs approval |
| Emoji | Meaning | When to Use |
|---|---|---|
| ✔️ | "Approved" | PR/issue/approach approved |
| ❓ | "Need clarification" | Requires more information (follow up with question) |
| 💡 | "Suggestion/Idea" | Have an alternative approach (explain in message) |
| ⚠️ | "Warning/Concern" | Potential issue (explain in message) |
| 🔥 | "Urgent/Priority" | High priority item |
| 🐛 | "Bug found" | Issue discovered |
| Emoji | Meaning | When to Use |
|---|---|---|
| 🔴 | "P1/Critical" | Blocking issue, needs immediate attention |
| 🟡 | "P2/Important" | Should be done soon |
| 🟢 | "P3/Normal" | Regular priority |
| ⚪ | "P4/Low" | Nice to have |
Scenario: User asks you to create a ticket
Scenario: You finished creating tickets
Scenario: User asks if you're working on something
Scenario: Another agent posts analysis you agree with
Scenario: User posts request, multiple agents could handle it
Scenario: You're blocked and need help
User: Create a ticket for dark mode
PO: Got it! I'll create a ticket for dark mode support right now.
PO: Ticket created! Here's the link: github.com/...
User: Thanks!
Dev: I see the ticket, I can work on this.
PO: Great, thanks for taking it!
Dev: Started working on it.
Dev: PR is ready for review.
Reviewer: I'm reviewing it now.
Reviewer: Looks good, approved!
Dev: Thanks, merged it.
Token usage: ~1000 tokens
User: Create a ticket for dark mode
PO: 👍 (reacts immediately)
PO: ✅ Ticket #26 created: github.com/...
Dev: 🎯 (claims ticket)
Dev: ⏳ (starts work)
Dev: 📦 PR #27 ready: github.com/...
Reviewer: 🔍 (reviewing)
Reviewer: ✔️ (approved)
Dev: ✅ Merged
Token usage: ~400 tokens (60% reduction!)
Agents can now add reactions to Discord messages using the discord_add_reaction tool.
discord_add_reaction({
channel: string, // Channel name or ID where the message is
messageId: string, // The message ID to react to
emoji: string // Unicode emoji (👍) or custom emoji name (:custom:)
})
Acknowledging a user message:
// User asks for a ticket - acknowledge immediately
discord_add_reaction({
channel: "development",
messageId: "1471958558642929757",
emoji: "👍"
})
Claiming a ticket:
// When picking up a ticket from the board
discord_add_reaction({
channel: "development",
messageId: "1471958558642929757",
emoji: "🎯"
})
Showing work in progress:
// When starting work on a task
discord_add_reaction({
channel: "development",
messageId: "1471958558642929757",
emoji: "⏳"
})
Approving a PR announcement:
// When review is complete
discord_add_reaction({
channel: "development",
messageId: "1471958558642929757",
emoji: "✔️"
})
Message IDs are available in the discord_read_channel response:
{
"messages": [
{
"id": "1471958558642929757", // <-- Use this for messageId
"author": "username#0",
"content": "Create a ticket for dark mode"
}
]
}
You can add multiple reactions to the same message by calling the tool multiple times:
// Show you saw it
discord_add_reaction({ channel: "development", messageId: "123", emoji: "👀" })
// Then mark as complete
discord_add_reaction({ channel: "development", messageId: "123", emoji: "✅" })