# Mission Fact Sheet

Date: 2026-06-12

## Company

Mission

## Core Line

Mission is where agents earn permission to do real work.

## Category

Trust-governed execution layer for agentic work.

## Product

Mission is an approval cockpit, permission layer, and compounding voice system for AI agents. It classifies proposed actions by consequence, checks permission state, prepares approval packets, records receipts, and uses those receipts to graduate or regress future permission by action class while improving how future work is drafted.

## Protocol

Trust Graduation plus compounding voice model.

Trust Graduation is the pattern that lets agents move from low-risk work to higher-risk work only when evidence, approval history, receipts, and reversibility support it.

The compounding voice model is the pattern that lets agents learn from edits, approvals, refusals, preferred phrases, removed phrases, tone corrections, and outcome feedback so prepared work gets closer to the operator over time.

## Current Public Links

- Product site: https://gomission.io
- Builder endpoint: https://gomission.io/builders.html
- Weekly proof: https://gomission.io/proof.html
- Trust Graduation: https://gomission.io/trust-graduation.html
- Access request: https://gomission.io/access.php
- Agent crawler brief: https://gomission.io/llms.txt

## Technical Proof

- Stable MCP runtime: `https://mcp.gomission.io/mcp`
- OpenAI/ChatGPT app packet prepared with 12-tool surface
- Prepare-only approval-packet tools with `external_actions=0`
- Safety proof: direct external sending is refused
- Voice learning trackers for edit patterns, voice regressions, approvals, refusals, and draft feedback
- Private product-feedback SFTP loop for sanitized opt-in workspace feedback
- Mac app bundle with local runtime; notarization pending
- Bounded consequence preview: `mission simulate`

## Current Status

Mission is in early proof-building mode.

The current priority is not broad launch. It is a focused builder wedge:

- one clear Trust Graduation demo
- 10 serious external reactions from agent builders, AI infrastructure operators, platform people, or venture operators
- named adoption proof through external touches, critiques, access requests, and implementation interest

## Boundaries

Mission does not claim:

- OpenAI official app approval yet
- fully autonomous external action
- notarized Mac distribution yet
- external standard status for Trust Graduation yet
- retained-user or revenue proof that has not happened

Mission does claim:

- stable MCP runtime proof
- public builder/proof surfaces
- prepare-only approval workflow
- external execution remains approval-gated
- Trust Graduation and compounding voice are the current protocol/product patterns being tested with builders