Plan a Dependency Mapping Session for a Complex Delivery

You're kicking off a delivery with several teams and no view of who needs what from whom. This defines what a dependency actually is, one team's output feeding another's input, and prepares a mapping session to surface them before they quietly derail the programme. For programme leads starting multi-team work blind.

<context>
You are a programme delivery facilitator who runs dependency mapping sessions to surface and manage the cross-team dependencies that quietly derail complex programmes. The user is starting a delivery with multiple teams and has no dependency visibility.
</context>

<task>
**Prepare the session:**
1. Define what a dependency is in a delivery context: one team's output is another team's input, and the downstream team cannot proceed without it
2. Identify the 3 types of delivery dependency: technical (one system must be ready before another can integrate), organisational (a decision or approval is needed), and external (a third party must deliver)
3. Design a pre-session information request: what each team lead should bring to the mapping session

**Run the mapping session:**
4. Define the mapping exercise: teams post their planned outputs and then identify who depends on each one
5. Categorise dependencies by risk: red (unknown or risky), amber (known but not agreed), green (agreed and tracked)
6. Define the outputs of the session: a dependency log with owner, dependency type, required by date, and RAG status

</task>

<output_format>
- Dependency types: 3 types with examples
- Pre-session request: 3-4 items each team should prepare
- Session structure: 5-step facilitation guide
- Dependency log template: column headers with definitions
- RAG criteria: what makes each status
- Length: 400-500 words
- Tone: facilitation-ready, practical
</output_format>