MCP server for emotional analysis, de-escalation, and strategic dialogue management in customer support
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"emotional-deescalation-mcp","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# Emotional De-escalation MCP Server v2 [Русская версия](README.ru.md) MCP server for **strategic dialogue management** and emotional de-escalation in customer support bots. The core problem: support bots get stuck in loops — asking for an order number 5 times, repeating the same apology, ignoring escalation signals. This server detects these patterns and tells the bot **what to do next** — change strategy, escalate to a human operator, or stop repeating itself. **Key capability: deterministic escalation detection.** The `strategy_suggest` tool identifies 7 problem patterns (repeated questions, legal threats, churn signals, requests for a human, emotional escalation, no progress, repeated contacts) and recommends concrete actions — including when to escalate to a human operator. Works without LLM calls, offline, instant. Based on: [Mistakes to Avoid When Developing Chatbots for User Support](https://medium.com/@ilyajob05/mistakes-to-avoid-when-developing-chatbots-for-user-support-5eefa21256ab) ## Strategic Dialogue Management The MVP covers 80% of problematic support scenarios **without any LLM calls**. The deterministic engine uses spaCy lemmatization and pattern matching to detect problems and recommend strategies: ### Detected Patterns | Pattern | What it detects | Escalation | |---------|----------------|------------| | **repeated_question** | Bot asked the same question 2+ times | After 4 repetitions | | **legal_threat** | User mentions lawsuits, regulators, lawyers | Immediate | | **human_request** | User explicitly asks for a live agent | Immediate | | **churn_signal** | User threatens to cancel, leave, or demands refund | After 1 more turn | | **emotion_escalation** | User's emotional intensity is increasing | After 2 more turns | | **no_progress** | Both sides repeating themselves, dialogue stuck | After 4 stagnant turns | | **repeated_contact** | User's Nth contact today (via metadata) | After 3rd contact | ### Escalation to Human Operator The …
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.