# agent loop detected re-asking answered compliance questions after failed save: how to fix it

The agent asks, gets an answer, fails to save it, and asks again. Verify each save with a read-back before moving on, so a failed save is caught immediately.

```text
agent loop detected re-asking answered compliance questions after failed save
```

## When to use this skill
- Agent asks the same compliance questions repeatedly
- Answers vanish between agent steps

## When not to use this skill
- Agent never asks the questions (a flow issue)
- Questions are asked once but answers are wrong (a quality issue)

## Tool compatibility
Agent questionnaire save verification. Related area: agent frameworks.

Related area: proposal automation and agent-assisted RFP workflows.

Cloud UIs change labels over time; if a button moved, search settings for the closest match.

## Before you start
- Sign in to the tool and keep one tab open so sessions stay clean
- Copy the exact error text and note the time it happened
- Set aside 15 to 20 minutes; most of these resolve in a single pass

## Quick diagnosis
Make sure this is the right skill before you invest the 15 minutes.
- If you are dealing with agent asks the same compliance questions repeatedly, start at step 1 below.
- If instead agent never asks the questions (a flow issue), stop here; you need a different skill.
- If neither quite matches, compare your screen against the verbatim error block at the top; that block is exactly what this skill covers.

## Fix it in steps
### 1. Read back after every save

```
After saving an answer, read the stored value back.
```
Expected: Expected: read-back matches what was saved.

### 2. Halt on save failure

```
If the read-back mismatches, stop instead of continuing.
```
Expected: Expected: failure surfaces at the question that broke.

### 3. Fix the save mechanism

```
Repair the underlying save: session, permissions, or field mapping.
```
Expected: Expected: saves persist reliably.

### 4. Resume from the failure point

```
Restart the questionnaire from the first unverified answer.
```
Expected: Expected: no question is asked twice.

## Confirm each step worked
Run through this checklist before moving on. If a check fails, redo that step instead of pushing ahead.
- Step 1 (read back after every save): Expected: expected: read-back matches what was saved.
- Step 2 (halt on save failure): Expected: expected: failure surfaces at the question that broke.
- Step 3 (fix the save mechanism): Expected: expected: saves persist reliably.
- Step 4 (resume from the failure point): Expected: expected: no question is asked twice.

## Quick recap
1. Read back after every save
2. Halt on save failure
3. Fix the save mechanism
4. Resume from the failure point

## Still stuck?
Gather three things: the exact error text, the time it happened, and what changed just before it broke.
Check the vendor status page for an active incident. Then open a support ticket with those three things;
it cuts the back-and-forth in half. For agent runs, attach the last 50 lines of the agent log too,
since the loop or timeout signature in the log usually names the cause.

## Variant phrasings
### agent re-asking compliance questions
Read-back verification fixes it. If this matches what you saw, the steps above apply as written.

### agent loop answered questions failed save
Halt on the first bad save. If this matches what you saw, the steps above apply as written.

## Why it happens
Silent save failures are the classic cause: the agent believes the answer persisted and moves on, then a later step finds it missing and re-asks. Read-back verification turns the silent failure into an immediate one.

## Edge cases
- Save failures often mean an expired session; re-authenticate first
- Batch saves hide which answer failed; save individually during debugging
- Log every save with its read-back result for the audit trail

## Provenance

Resolved from the public thread: https://vectle.com/posts/pst_UevFNkDqlC-SZRotBmCbTA