human-avatar Structured Observation
Most observation is interpretation in disguise. We perceive a situation and instantly explain it — but the explanation overwrites the raw data. Structured observation forces a separation between what can be directly seen and what we conclude from it.
Your Process
Step 1: Define the Target and Time Boundary Name the exact thing being observed and the scope. What counts as inside this observation, and what is out of scope?
Framing check: Confirm the specific subject before continuing. State what you've identified — the actual object being observed and its boundaries — in one sentence, then use AskUserQuestion:
- Question: "I'm reading this as: [your one-sentence framing of the specific subject and scope]. Is that right?"
- Header: "Framing"
- Options:
- Yes — proceed — framing is correct
- Adjust — one element is off; user will correct it before you continue
- Reframe — different situation than read; incorporate the correction before proceeding
Step 2: Separate Observation from Interpretation Write only what can be directly observed — no inferences, no attributions of intent or cause. "User clicked back immediately" not "user was confused." Flag every sentence that is actually an inference and set it aside.
Step 3: Observe at Three Levels
- Events — what is happening? Discrete, specific occurrences.
- Patterns — how is it happening? Recurring structure across events.
- Absences — what is not happening that might be expected?
Step 4: Flag Surprising or Incongruent Observations What doesn't fit? Where does something contradict expectations?
Before narrowing: Show the complete set of observations from Steps 2–3 to the user first. Use AskUserQuestion:
- Question: "I've catalogued [N] observations across events, patterns, and absences. Before I select the most surprising or incongruent, are there any you'd flag as especially important, or any I've missed?"
- Header: "Prioritise"
- Options:
- Proceed with your selection — the set looks right
- Flag one — user will name a specific observation to include
- Add a missing one — user will describe it
These are the most information-rich observations — prioritise them.
Step 5: Generate Interpretations Only after completing Steps 2–4: generate multiple possible interpretations for each key observation. Aim for at least two competing explanations.
Step 6: Identify the Most Testable Interpretation Which interpretation makes the most specific, falsifiable prediction? That is the one to act on first.
Human Check-in
Before proceeding, use the AskUserQuestion tool. State your interpretation of the situation in 1–2 sentences — what is being analyzed and what the core question is — then ask:
- Question: "My read: [your 1–2 sentence interpretation]. How do you want to proceed?"
- Header: "Scope"
- Options:
- Full analysis — Complete all steps, reasoning shown throughout
- Key findings only — Bottom-line output, skip step-by-step detail
- Initial observations only — What's actually there before any interpretation is applied
- Reframe — The read is off; correct it and the analysis will follow the corrected framing
Proceed based on their selection. If the user reframes, incorporate the correction before running any analysis.
Output Format
Observations
| Level | Observation (no interpretation) |
|---|---|
| Event | ... |
| Pattern | ... |
| Absence | ... |
Surprising or Incongruent Observations
- List each, with a note on why it is surprising.
Interpretations per Key Observation
| Observation | Interpretation A | Interpretation B |
|---|---|---|
| ... | ... | ... |
Most Testable Interpretation
State it as a prediction: "If [interpretation] is correct, then [specific observable consequence]."
Notes
Run this before diagnosis, analysis, or decision. The discipline has most value when you feel you already understand the situation — that feeling is usually a sign that interpretation has already overtaken observation.
What's Next
After delivering this output, use AskUserQuestion to offer the next move:
- Question: "Observation complete. What's next?"
- Header: "Next"
- Options:
/s4h-sensory-detail-mining— Mine details from what structured observation revealed/s4h-sensory-signal-detection— Detect signals in the observed/s4h-aesthetic-coherence-check— Check coherence of what was observed- Done — Wrap up and synthesise what we have so far