Integrity Analysis v0.1
Integrity Forensics Generator
Examine distortion, instability, and evidentiary weakness without overstating the case.
Use this generator to assess a passage, translation, teaching, or claim with structured restraint. It separates observed evidence from inference, assigns a bounded evidence grade, and keeps ambiguity visible when the material remains inconclusive.
Operating Posture
Disciplined inquiry before verdict language.
Separate observation from inference.
Let ambiguity remain visible when evidence is thin.
Treat anomaly as a signal to inspect, not proof to declare.
This tool supports disciplined inquiry, not reckless accusation.
It is not a lie detector, oracle, prosecutor, or final spiritual judge.
BOUNDED OUTPUT
What this generator returns
A structured assessment with distortion classes, signal buckets, evidence / inference separation, a bounded conclusion, and reflection prompts.
Three focused modes
Scripture review, translation review, and claim review only. The v1 scope stays narrow on purpose.
Evidence-aware output
The response separates directly observed evidence from reasonable inference, unsupported concerns, and possible user bias.
Generator-family behavior
Save local drafts, reload recent runs, and export structured results without making raw markdown the primary UI.
Guided Intake
Integrity Forensics Generator
Choose the assessment mode, define the object of analysis precisely, and provide only the material needed to test the concern.
Example seeds
Start with the example that matches your current mode, then adapt it to your case.
Assessment Output
Structured result cards
No assessment yet
Output Preview
Your assessment will land here as a structured forensic stack.
- Distortion Profile and Evidence Grade surface first.
- Signal cards appear only when the model has something relevant to say.
- Evidence / Inference Split always stays explicit.
What Integrity Forensics Generator does
Integrity Forensics organizes a narrow, evidence-aware inquiry into a text, translation, interpretation, transmission question, or claim. An anomaly is not proof, and the tool is not a lie detector.
Problem class
A bounded concern requiring explicit separation of observation, inference, unsupported concern, and uncertainty.
Input contract
The object of analysis, concern, supplied material, and selected inquiry mode.
When to use
- Reviewing a supplied claim or material.
- Clarifying what evidence would strengthen an inquiry.
When not to use
- Declaring corruption, guilt, sincerity, or institutional intent.
- Replacing qualified textual, historical, translation, or legal expertise.
Method stages
- Identify the object and mode.
- Separate evidence from inference.
- Grade the available signal conservatively.
- State what would strengthen review.
Output contract
A bounded conclusion, evidence grade, and open questions.
Interpretation guidance
Read grades as limits on confidence, not as accusations.
Worked example
Synthetic example: A synthetic reader notices a translation difference; the output distinguishes the wording supplied from an inference about cause and identifies evidence needed before a claim.
Unresolved unknowns and limits
The tool cannot infer sincerity, prove corruption, or establish manuscript authority. Accusatory claims remain outside the tool’s authority.
Privacy and storage
Use only information you can responsibly provide. Storage behavior is shown by the generator after a run; do not treat this page as a guarantee of a particular storage path.
Related questions
- What facts, assumptions, and unknowns need to stay separate?
- What would make the next step more responsible?
Related generators
MandalaStacks is an applied system, not a final authority. You retain responsibility for decisions and actions. Reviewed 2026-07-30.