# a11y patch agent failed on dynamically injected aria-live region - how to fix it

## TL;DR

Patch the live region's creator, not the region: dynamically injected aria-live regions get wiped on the next render, so the agent must fix the component or script that creates them. Re-applying attributes to a node that will be replaced is futile. One line of why: the region is ephemeral, its creator is persistent, and patches must target what persists.

## The error, verbatim

```text
RemediationAgentError: live region patch lost on rerender
    patched: aria-live=polite on #toast-region
    after_rerender: node replaced, attributes gone
    framework: toast library recreates region per message

```

## Fix it step by step

### Step 1: Reproduce the lost patch

```bash
node agent/remediate.js --rule aria-live | rg -i 'lost|rerender|replaced' | head -5
```

Expected: The patch applies then vanishes on the next render.

### Step 2: Find the region creator

```bash
rg -n 'aria-live|toast' src --glob '*.{jsx,tsx,js}' | head -10
```

Expected: Shows the toast library or component creating the region.

### Step 3: Patch the creator

```bash
rg -n 'live' agent/patches/aria-live.js | head -10
```

Expected: Change the patch target from the DOM node to the creator's config or source.

### Step 4: Re-run remediation

```bash
node agent/remediate.js --rule aria-live | tail -4
```

Expected: Region is created correctly and survives rerenders.

### Step 5: Add a regression probe

```bash
node agent/run-audit.js --smoke | tail -3
```

Expected: Smoke run passes; schedule it so the breakdown is caught if it ever regresses.

## When to use this skill

- You run an agent that scans UIs for accessibility and it hits this breakdown
- The agent's scan loop stalls, crashes, or loops on this exact failure
- You are hardening an audit agent's error handling for production scans

## When NOT to use this skill

- A human runs the scan manually and it works, this is agent-harness failure handling
- The scan completes and only reports violations, use the rule-specific skills

## Compatibility

Remediation agent with source patch ability. Toast and notification libraries are the usual creators. Pin the tool version in the lockfile so scans stay reproducible across machines.

## Variant phrasings

### agent live region patch overwritten

Same loss, same creator-level fix.

### aria-live injected dynamically remediation

Practitioner phrasing.

### the breakdown hits other routes too

Agent failure modes are systemic; apply the hardening to every route the agent covers, not just the one that failed.

## Why it happens

Toast libraries and notification systems create aria-live regions dynamically per message, replacing the node each time. An agent patching attributes on the live node watches its work evaporate on the next toast. The durable patch configures the creator: the library's options, the component's props, or the template that emits the region. This is the same lesson as the label rerender loop: patch what persists, not what renders. Agent breakdowns are systemic: the same failure mode will hit every route, page, or run the agent touches. Harden the harness once (timeouts, loop detection, verification gates) instead of patching per page, and keep breakdown telemetry separate from violation counts.

## Edge cases

- Some libraries expose no config for the live region, then the fix is a wrapper component or a library swap.
- Assertive vs polite is a product decision, the agent should preserve the intended politeness, not upgrade it.
- Verify across two toast cycles, one render is not proof the patch sticks.
- Log breakdowns separately from violations in agent telemetry; mixing them hides whether the agent itself is getting more reliable.

## Provenance

Resolved from the public thread: https://vectle.com/posts/pst_nxdFuehq2d8RSnWK5GG8IQ