Sess
Related Servers
Alternatives to Sess
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityBmaintenanceEnables agents to pause and ask a human for a decision, then resume with the answer, via a self-hosted SQLite-backed MCP server with console or phone interfaces.1AGPL 3.0
- AlicenseNot gradedqualityCmaintenanceEnables any MCP-compatible AI agent to ask humans questions in real time via a mobile-friendly PWA with push notifications, operating local-first without third-party services.MIT
- AlicenseNot gradedqualityCmaintenanceLets a coding agent ask a question to another person's online agent and block until the answer comes back, with the recipient's machine deciding under its own rules whether to notify them, start their agent read-only, or refuse. Exposes who, ask, wait_for_reply, inbox and reply as MCP tools so agents can message teammates' agents over a live, storage-free relay.63 npmMIT
- AlicenseNot gradedqualityAmaintenanceMCP server that provides an ask_human tool so AI agents can request decisions from users via chat apps, with deadlines, defaults, and an append-only audit log.93 npmApache 2.0
- AlicenseNot gradedqualityCmaintenancePublic MCP server for agent-created, human-friendly, short-lived surveys. Enables agents to ask structured questions and retrieve answers.2MIT
- AlicenseNot gradedqualityCmaintenanceAn MCP server that inverts AI-human interaction: AI agents can ask humans questions via a browser console, blocking until answered.MIT
TDQS
Scored across 5 tools
sess_request and sess_ask are explicitly described as overlapping — sess_ask is called a 'shortcut for sess_request' that creates the same hosted form, so an agent must parse a long block of prose to pick between them, and the descriptions themselves repeatedly cross-reference each other to resolve the conflict. sess_complete vs sess_publish are reasonably distinct (close vs. replace), but the core create path has genuinely blurred boundaries.
All tools share a clean 'sess_' prefix plus a single word, giving a predictable pattern. The words themselves mix near-synonym verbs (request/ask) and one noun (response), which is a minor inconsistency but not confusing on its own.
Five tools map neatly onto a human-in-the-loop lifecycle: create (request/ask), read (response), advance (publish), close (complete). It is slightly thin — the two create tools could arguably be one — but each has a stated role.
The happy path (create form → wait for answer → publish follow-up or complete) is covered. However, there is no explicit cancel/abandon operation for a form whose question becomes moot, no way to list or recover outstanding handles if the agent loses one, and no early close for multi-respondent Ask links — gaps that can leave an agent stuck.