DecisionPing
Human-in-the-loop infrastructure for AI agents: explicit approvals, choices, and text answers through MCP.
Date not announced yet
For whom: Developers and teams building autonomous AI agents, coding agents, operations workflows, release automation, and other systems that sometimes need an explicit human decision.
Already done: The hosted API, administration dashboard, remote MCP server, Telegram and WWChat delivery, structured responses, discovery documents, Privacy Policy, and Terms of Use are live.
After joining: Beta users can connect working agents today, influence the roadmap, and help shape native clients, additional delivery channels, and bring-your-own-bot support.
About
DecisionPing is human-in-the-loop infrastructure for autonomous AI agents. When an agent reaches a decision it should not make alone, it can notify a human or request confirmation, a choice, or free text through a remote MCP server. Requests are delivered to Telegram and WWChat, and the agent resumes only after an explicit structured response.
The beta includes a production API and dashboard, project-scoped revocable keys, durable request IDs, idempotent retries, expiration and cancellation, delivery history, standard MCP tool annotations, llms.txt discovery, and five MCP tools: notify, ask_user, get_request, wait_for_response, and cancel_request.
Connect any compatible agent using decisionping.com/connect. Native clients and bring-your-own bots are next.
Ideas (0)
No ideas yet. Suggest the first one — the team will appreciate the feedback.
Updates (0)
No updates yet. The author hasn't posted any news.
Discussion (0)
Sign in to leave a comment