- Home
- Skills
- Agents & Orchestration
- AMDA Decision Router
AMDA Decision Router
Classify any AI output by whether a human needs to review it, and why.
Free
Works with the AI Agents you already use
AMDA Decision Router
Example session with this skill installed
Our AI agent drafts the weekly customer newsletter and schedules it to send automatically. Does this need human review?
- Read your context and instructions
- Compiled the amda decision router
Decision classified
Decision: AI agent drafts and auto-sends the weekly customer newsletter
| Question | Answer |
|---|---|
| Reversible? | No — once sent, it's in 12,000 inboxes |
| Exposure? | High — external, customer-facing, brand voice |
| Frequency? | Weekly — moderate |
→ Zone 2: Collaborative
Governance rule: AI drafts the newsletter. A human reviews and approves before send. The AI must not hold the send permission.
If you classify one zone too low (Zone 1): A compliance exception accumulates quietly. Discovery is usually an incident, not an audit.
If you classify one zone too high (Zone 3): A named senior person becomes the bottleneck for a weekly task. Time to publish grows until the tool gets abandoned.
The key distinction this surfaces: Drafting is Zone 1 work. Sending is Zone 2. When a workflow spans zones, govern at the boundary — not at the highest zone across the whole workflow. That split is where most of the adoption is won.
Connects securely to your tools. The creator never sees your data.
About this skill
Most AI governance fails in one of two directions. Either every output gets a human review and adoption collapses, or nothing gets reviewed and something blows up.
The AMDA Decision Router classifies any marketing or GTM decision into one of three zones (AI-Primary, Collaborative, Human-Primary) using three questions about reversibility, exposure, and frequency. It returns the governance rule for that zone, the failure mode if you classify too low, and the failure mode if you classify too high.
Built on the Augmented Marketing Decision Architecture by Kuber Sharma. Adopted into the claude-skills open-source repo (28.5k stars, v2.13.0, Oct 2026). First-published from 16 enterprise product launches at Tableau, a Salesforce company.
Free. CC BY-ND 4.0. Full specification: kubersharma.com/frameworks/amda
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Based on observational studies in marketing contexts, not a controlled experiment. Does not provide legal or compliance guarantees. High-frequency estimates are context-specific.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 20 days ago
- Free with an Agensi account
Permissions
No special permissions declared or detected