Mission is a human-gated AI operating layer. These terms describe the practical boundaries for using the product, pilots, local workspaces, agent APIs, and approval-gated workflows.
Mission may prepare, rank, summarize, draft, route, and evaluate work. External actions such as sending messages, publishing, spending, changing accounts, or making commitments require human approval unless a written pilot agreement says otherwise.
You are responsible for reviewing outputs before use and deciding whether a prepared action is appropriate.
Do not use Mission for spam, deceptive outreach, credential harvesting, surveillance, harassment, illegal activity, or fully unsupervised high-consequence decisions. Do not bypass approval gates, safety prompts, rate limits, or local workspace controls.
Founding Partner installs and pilots are founder-led services unless otherwise agreed in writing. Scope, pricing, duration, included connectors, and support expectations should be confirmed before real production use.
Mission produces recommendations from available context and may be incomplete or wrong. Receipts, source evidence, edits, rejections, and outcomes are part of the product because prepared work is not the same as executed work.
For formal commercial terms, enterprise requirements, or legal review, contact founders@phenomenalabs.com.