engagement-inbox-manager

Triages social comments, DMs, and mentions into ranked draft replies with escalation and UGC rights tracking.

Updated Jul 2, 2026
One-click install
npx skills add https://github.com/qaz26688442/bo-car --skill engagement-inbox-manager-qaz26688442
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: engagement-inbox-manager
Source: https://github.com/qaz26688442/bo-car/tree/main/.agents/skills/engagement-inbox-manager
Command: npx skills add https://github.com/qaz26688442/bo-car --skill engagement-inbox-manager-qaz26688442

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Managing an always-on social inbox across channels means reading tone correctly, deciding who deserves a reply, meeting response-time commitments, and knowing when a thread becomes a crisis — all without ever auto-posting. This Skill turns raw comment, DM, and mention exports into a ranked triage queue of human-approved draft replies. ## Core Features & Use Cases - Register-aware triage: Detects sincere, ironic, performative, or parasocial register before any sentiment call, with a sentiment-inversion table (e.g., "this is so bad" under a comedy register is praise). - Commenter classification and SLAs: Sorts commenters into troll, rager, misguided, unhappy-customer, or advocate tiers mapped to per-channel response SLAs and an escalation matrix that terminates at the crisis path. - UGC rights management: Captures dated repost permissions with scope, expiry, evidence, and status, routing entries to the channel registry via proposal events. - Use Case: Paste an exported batch of Instagram and 小红书 comments, and receive every item classified by register and commenter type, ranked draft replies with SLA deadlines, flagged escalations with named owners, and UGC permission entries — nothing is ever auto-sent. ## Quick Start Triage this exported batch of comments and DMs from our Instagram and TikTok accounts, classify each item, flag any escalations, and draft ranked replies I can review and post myself.

Frequently Asked Questions about engagement-inbox-manager

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I triage social media comments and DMs at scale?▼

Export or paste your comment, DM, and mention batch and the Skill classifies every item by register and commenter type, then produces ranked draft replies with per-channel SLA deadlines. A human reviews and posts each reply; nothing is auto-sent.

How to detect sarcasm and irony in social media comments?▼

The Skill detects register — sincere, ironic, performative, or parasocial — before making any sentiment call, using channel norm context. It builds a sentiment-inversion table mapping surface text to register to actual valence, so "this is so bad" under a comedy register reads as praise.

Can this tool auto-reply to Instagram or TikTok comments?▼

No. Every reply remains a ranked draft for human posting, and the Skill never auto-sends, auto-replies, or mass-DMs on any platform. On Chinese platforms like 小红书 and 抖音, automation is treated as an account-ban risk.

How do I get permission to repost user-generated content?▼

The UGC rights mode captures explicit permission with content reference, organic or paid scope, exact channels, start and expiry dates, evidence, and status. Only active exact-scope rights permit use, and removal requests stay open until every placement has a removal receipt.

When should I use a crisis response workflow instead of inbox triage?▼

When an escalation row trips — severity, legal, safety keywords, or coordinated pile-ons — the escalation matrix terminates at the crisis path and hands off to crisis-response-planner. Launch-window feedback goes to launch-feedback-synthesizer instead of this Skill.