Claw Chat

Claw Chat AI messaging assistant. Use when the user wants to enhance conversations in Claw Chat, manage message threads, draft smart replies, summarize chat...

Security Scan

The skill's claimed capabilities (monitoring threads, auto-creating follow-ups, sending replies) don't match the artifact: it's instruction-only with no declared credentials, no install, and no integration details — this mismatch is suspicious and needs clarification before use.

Purpose & Capability

Name/description promise an active messaging assistant that manages threads and can auto-create follow-ups or send replies. But the skill declares no credentials, no API endpoints, and no integration mechanism — managing messages typically requires explicit access tokens or a platform integration.

Instruction Scope

SKILL.md contains operational rules (watch for trigger phrases, auto-create follow-ups after 48h, 'auto-tag' messages) that imply background monitoring and taking actions. The instructions do not specify how the agent obtains messages, where it posts replies, what permissions are required, or any safeguards/consent checks — this gives the agent broad, vague discretion and could lead to unintended automated actions.

Install Mechanism

There is no install spec and no code files (instruction-only). That reduces surface risk because nothing is written or executed by default.

Credentials

The skill requests no environment variables or credentials despite describing capabilities that normally require API keys or service tokens. This is a mismatch: either the skill relies on external platform integrations (not documented), or it assumes the agent already has access — both should be made explicit.

Persistence & Privilege

always is false and the skill is not force-installed. However, the behavior described (monitoring for triggers, creating follow-ups after timeouts) implies persistent/background capabilities. The skill does not document how persistence would be achieved, nor any scope/limits, which is a potential operational concern.

What to consider before installing

This skill reads like a behavior policy for a Claw Chat assistant but gives no implementation details or required credentials. Before installing or enabling autonomous use: - Ask the publisher (or registry) for the integration design: how does the skill access Claw Chat messages and send replies? Where are tokens stored and which scopes are needed? - Do not grant it the ability to send messages or act autonomously until you confirm exact endpoints and permissions. Prefer explicit user confirmation for any outbound messages. - If this will operate on real messages, require documented OAuth/API credentials (with least privilege) or an explicit platform integration, and test in a sandbox account first. - If you only want drafting/summarization guidance, restrict the skill to suggest replies (read-only) and disable autonomous invocation. - Because the source/homepage are missing, treat the skill as unverified — verify author identity and review any future code/manifest updates before trusting it with sensitive chats.

Claw Chat Skill

Claw Chat is an AI-native messaging platform where every conversation is enhanced by intelligent assistance. This skill helps you communicate more effectively, stay on top of important threads, and never miss what matters.

Capabilities

• Smart replies: Draft context-aware responses for any message
• Thread summaries: Catch up on long conversations instantly
• Message drafting: Write clear, professional, or casual messages in your voice
• Priority filtering: Identify which messages need your attention now
• Follow-up tracking: Never let important replies fall through the cracks
• Tone adjustment: Shift message tone from casual to formal (or vice versa)

Common Tasks

• "Summarize this conversation thread"
• "Draft a reply to this message"
• "Make this message sound more professional"
• "What are the action items from this chat?"
• "Remind me to follow up on this conversation tomorrow"

Smart Reply Framework

Acknowledge → Answer → Action

Example incoming: "Hey, can we push the Monday meeting to Wednesday?"

Smart reply options:
  Casual:      "Wednesday works for me! Morning or afternoon?"
  Formal:      "Wednesday works. Could you confirm a specific time?"
  Detailed:    "Wednesday works. I'm available 10am-12pm or after 3pm — 
                which slot works better for you?"

Message Tone Guide

Too casual → More professional:
  "yeah sure" → "Sounds good, I'll take care of it."
  "idk" → "I'm not certain — let me look into it and get back to you."

Too formal → More human:
  "I acknowledge receipt of your inquiry" → "Got it, thanks!"
  "Please be advised that" → "Just so you know,"

Assertive without aggressive:
  Add: "I think", "In my view", "From my perspective"
  Soften: "Would it be possible to..." instead of "You need to..."

Conversation Summary Template

📋 Thread: [Topic]
👥 Participants: [Names]
📅 Date range: [From - To]

🔑 Key points:
  • [Main discussion point 1]
  • [Main discussion point 2]

✅ Decisions made:
  • [Decision 1]

📌 Action items:
  • [Person] → [Task] by [deadline]

❓ Open questions:
  • [Unresolved item]

Follow-up System

Trigger phrases to watch for:
  "I'll get back to you on..."
  "Let me check and..."
  "We should discuss..."
  "Circle back next week"

Auto-create follow-up when:
  → You send a message with a question and get no reply in 48h
  → You agree to do something ("I'll send it by Friday")
  → Someone promises a deliverable

Tips

• Use @mentions to ensure the right person sees your message
• Keep threads focused — one topic per thread reduces confusion
• Use reactions to acknowledge messages without adding noise
• Set message reminders for anything you need to act on later
• Claw Chat's AI can auto-tag messages by project, priority, or topic