Hey PH! I'm the maker of EQMO. I kept hitting the same problem: AI agents are great at producing documents — plans, policies, contracts, runbooks — but someone still has to actually read and sign off on them. Agents had no way to wait for a human "yes." EQMO. closes that loop: — Agents submit documents via MCP or a REST API (bearer key, 5 scopes) and long-poll for the decision. — Humans get an approval email with a one-time link — no account needed. Highlight any text to comment; comments re-anchor when the agent submits a new version. — Approval chains (sequential or unanimous, 1–5 approvers), immutable versions with word-level diffs, and an append-only, hash-chained audit log. Everything an approver touches is a dashboard; everything an agent touches is an API. If your agent can call MCP, it can use EQMO — there's an MCP server at mcp.eqmo.com and a skill it can install. Free plan: 1 project, 50 docs/month. Paid from $19/mo. Ask me anything — especially about the "approved with comments, mandatory annex in the next version" flow, which I'm most proud of. Thanks for reading.

EQMO.

Human approval for what your AI agent writes

Human approval for what your AI agent writes