> The first act draws the organisation's knowledge as a graph. Pick a real question, trace down what knowledge answering it depends on, and name who holds each piece. It maps the human network the organisation actually runs on - who you go to - while the question of whether any of it is *written down* is held back for the next act. Connections first; storage and state come later.
^purpose
1. We pick a concrete prompt and mark who raises it - a person or an [[AI agent|agent]]
2. We trace its dependency cone downward - the knowledge it needs, and what that needs - as a DAG
3. We name the holder of each piece: the real who-you-ask network, not the org chart
4. We re-walk the cone as an agent - which needs every piece written and reachable, where a person just asks a colleague; the gap between the two is the problem the whole workshop exists to close
^plan
The act draws on knowledge-audit practice and organisational network analysis. Two distinctions set it apart: AI agents are treated as first-class knowledge consumers alongside people, and the orientation is investigative - toward what a question actually depends on, not a catalogue of what exists. Everything is laid down uncoloured: this act records structure, the next act judges state.
**You leave with:**
- A traced knowledge graph for at least one real prompt - the question, its dependency cone, and the people each piece rests on
- The bus factors made visible - a piece many prompts depend on, held by one person
- The human-vs-agent gap named - what the organisation navigates by asking people, an agent (or a newcomer after a reorg) cannot
^outcomes
## Examples traced here
1. [[Second brain example - personal notes|Second brain (personal notes)]]
![[Second brain example - personal notes#^trail-line]]
2. [[Shared second brain example - ADR log|Shared second brain (ADR log)]]
![[Shared second brain example - ADR log#^trail-line]]
3. [[Shared second brain example - Expertise Radar|Shared second brain (Expertise Radar)]]
![[Shared second brain example - Expertise Radar#^trail-line]]
4. [[Shared second brain example - Experts Hub|Shared second brain (Experts Hub)]]
![[Shared second brain example - Experts Hub#^trail-line]]