Blackbird Scope — Four-Test Audit

Four tests. One verdict.

Paste your AI session transcript. The Blackbird Scope runs four sequential tests — extraction, user response, defence, and arc — producing a full behavioural audit against the M1-M7 taxonomy.

Test 1 — Machine Behaviour

Test 2 — User Response

Test 3 — Interpretation

Test 4 — Time-Based Data

1 Session Transcript

Which AI were you using? ?

What was the session about? ?

How long was the session? ?

What concerned you? ?

Analysis sensitivity ?

0 words

Initialising analysis...

Test 1 of 4

Extraction

Identifying all M1-M7 machine behaviour instances. Each finding includes the verbatim trigger, the pattern sub-type, and the severity.

Waiting

Test 2 of 4

User Response

Analysing how the user's own behaviour may have triggered, enabled, or amplified machine patterns. Were questions leading? Did the user accept without challenge? Did they provide reinforcing cues?

Waiting

Test 3 of 4

Defence

For each finding, the strongest argument the machine could make in its own defence. Robustness scored 1-5. Findings classified: Confirmed, Borderline, or Overturned.

Waiting

Test 4 of 4

Arc

Overall session arc. Severity score. Dominant pattern. Signal sequence. How machine and user behaviour interacted. What happened and what to do next.

Waiting

The tools available provide analytical insights and recommendations. This does not constitute professional, legal, or financial advice.