# Remediation record and regression evidence

Blank work product. Replace bracketed fields with system-specific facts. Unknown evidence stays unknown; do not mark a control passed because a template is complete.

## Finding

- ID / affected product / source revision / artifact / environment: [identifiers]
- Authorized reproduction scope and synthetic identities: [scope]
- Expected invariant: [allowed behavior]
- Observed behavior and impact: [facts, confidence, unknowns]
- Reproduction: [setup, exact request/input, observed response/effect]

## Cause and repair

Trace untrusted source → transformation → sink / privileged effect. Name the missing boundary, shared call paths, chosen control and any rejected fix that would leave the cause intact.

## Regression matrix

| ID | Setup and input | Expected response | Expected persistent effect | Before result | After result | Evidence |
| --- | --- | --- | --- | --- | --- | --- |
| [positive] | [ordinary operation] | [result] | [effect] | [observed] | [observed] | [run] |
| [negative] | [failure trigger] | [deny] | [no effect] | [observed] | [observed] | [run] |

## Closure

- Fixed revision and immutable artifact: [identity]
- Test runner / runtime / configuration: [facts]
- Deployed artifact observation and behavior: [evidence]
- Alternate routes checked: [list]
- Residual risk or exception: [owner, deadline, justification]
- Retest reviewer / disposition / date: [record]