---
name: config-audit:drift
description: Compare current configuration against a saved baseline — shows new, resolved, and changed findings
argument-hint: "[path] [--baseline name] [--save]"
allowed-tools: Read, Write, Glob, Grep, Bash
model: sonnet
---

# Config-Audit: Drift Detection

Compare current configuration against a saved baseline to see what changed.

## Arguments

- `$ARGUMENTS` may contain:
  - A target path (default: current working directory)
  - `--save`: Save current state as baseline
  - `--baseline <name>`: Compare against a specific named baseline (default: "default")
  - `--raw`: Pass-through to the scanner; produces v5.0.0 verbatim diff output (bypasses the humanizer). Use when piping into v5.0.0-baseline diff tooling that depends on byte-stable output.

## Implementation

### Save a baseline

If `--save` is present:

Tell the user: **"Saving current configuration as baseline..."**

```bash
RAW_FLAG=""
if echo "$ARGUMENTS" | grep -q -- "--raw"; then RAW_FLAG="--raw"; fi
node ${CLAUDE_PLUGIN_ROOT}/scanners/drift-cli.mjs <path> --save --name <baseline-name> $RAW_FLAG 2>/dev/null
```

Read stdout for confirmation. Tell the user:

```markdown
### Baseline Saved

Captured current state as baseline "{name}".
Run `/config-audit drift` anytime to see what changed since this point.
```

### Compare against baseline

Without `--save`:

Tell the user: **"Comparing current configuration against baseline..."**

```bash
RAW_FLAG=""
if echo "$ARGUMENTS" | grep -q -- "--raw"; then RAW_FLAG="--raw"; fi
node ${CLAUDE_PLUGIN_ROOT}/scanners/drift-cli.mjs <path> --baseline <name> $RAW_FLAG 2>/dev/null
```

Read stdout. In default mode the diff sections are humanized — finding titles, descriptions, and recommendations have already been replaced with plain-language equivalents. New/resolved/changed finding lists carry `userImpactCategory`, `userActionLanguage`, and `relevanceContext` so you can group and prioritize without re-deriving severity prose. If `--raw` was passed, the v5.0.0 diff is verbatim — present it in a code block as-is.

If baseline not found, tell the user:

```
No baseline found. Save one first with:
  /config-audit drift --save
```

Otherwise, parse and present the drift report. Use the Read tool on the captured stdout (or pipe it into a tmpfile first if you prefer):

```markdown
### Configuration Drift

**Trend:** {Improving|Degrading|Stable}
**Score:** {before} → {after} ({+/-delta} points)

{If new findings:}
#### New Issues ({count})
| ID | Action | Description |
|----|--------|-------------|
| {id} | {userActionLanguage — "Fix this now", "Fix soon", etc.} | {humanized title} |

{If resolved findings:}
#### Resolved ({count})
| ID | Description |
|----|-------------|
| {id} | {humanized title} |

{If area changes:}
#### Area Changes
| Area | Before | After | Change |
|------|--------|-------|--------|
| ... | ... | ... | ... |
```

When iterating new/resolved findings, prefer `userActionLanguage` over raw `severity` for the "Action" column — the humanizer already mapped severity to plain-language phrasing, and surfacing it consistently keeps the toolchain coherent. Mention `relevanceContext` when it isn't `affects-everyone` (the user wants to know if a fix touches shared config or just their machine).

### List baselines

If `$ARGUMENTS` contains `--list`:

```bash
node ${CLAUDE_PLUGIN_ROOT}/scanners/drift-cli.mjs --list 2>/dev/null
```

### What's next

After viewing drift:
- `/config-audit fix` — Auto-fix new findings
- `/config-audit posture` — Full posture assessment
- `/config-audit drift --save` — Update the baseline to current state