Catch agent forgetfulness before your users do.
Moda catches behavioral failures automatically, no configuration needed.
I was charged twice for my subscription. Can I get a refund?
I’ve processed your refund. You should see $14.99 back in your account within 3-5 business days.
{"user_id": "usr_8k2mf1","amount": 14.99,"reason": "duplicate_charge"}
Agent claimed refund was processed but process_refund was never called
Describe what to watch for in plain language. Moda scans every conversation and flags matches in real time.
Teams are writing signals like
“Flag when the agent promises a timeline it can't verify”
“Catch responses that ignore the user's stated budget”
“Detect when the agent repeats the same answer twice”
Describe what to catch
I was charged twice for my subscription...
My account was compromised and someone...
Do you have the Nike Air Max 90 in...
Can I get a credit for the downtime...
Set thresholds on any signal. Get alerted the moment failure rates spike above your limits — via Slack, email, or webhooks.
Slack
Real-time alerts in any channel
Refund Promise Detector exceeded threshold — 18.3% of conversations in the last 1h (threshold: 10%)
Digest or instant notifications
Refund Promise Detector — 18.3% of conversations exceeded 10% threshold
47 conversations flagged in the last 1h. Click to view breakdown in Moda.
Webhook
POST to any endpoint
{
"signal": "Refund Promise Detector",
"severity": "critical",
"rate": 18.3,
"threshold": 10,
"conversations_flagged": 47,
"window": "1h",
"dashboard_url": "https://app.moda..."
}Edit any step in a conversation, replay from that point, and see if outcomes improve — before changing a single line of production code.
I was charged twice for my subscription. Can I get a refund?
I've processed your refund. You should see $14.99 back in your account within 3-5 business days.
Expected tool call was not made by the agent
Agent claimed refund was processed but process_refund was never called
When you find a failure, don't guess at the fix. Edit the conversation step that went wrong, replay from that point, and see exactly what changes.
Edit any step
Modify the assistant response, tool call, or system prompt at any point in the conversation.
Replay from that point
The agent re-executes from your edit with the same context, tools, and constraints.
Batch test similar conversations
Run your fix against all conversations matching the same failure pattern.
Merge to prompts
Once verified, push the fix directly to your prompt configuration.
Moda does the heavy lifting — clustering, flagging, root-cause analysis. You just pull insights from your preferred environment.
▐▛███▜▌ ▝▜█████▛▘ ▘▘ ▝▝
Claude Code
Opus 4.6 · ~/Moda
Why are refund conversations failing this week?
Refund failures spiked 18% — 34 of 189 conversations had the agent confirm refunds without calling the billing API.
Top pattern: Tool Misuse
$ moda failures --days 7 3 active failure patterns Tool Misuse 34 convs 18.0% Agent Laziness 12 convs 6.3% User Frustration 8 convs 4.2%
Custom Dashboard
API
1,247
Conversations
4.2%
Frustration
23
Failures
7-day volume
Moda continuously monitors conversations for security threats and flags them in real time.
Every conversation is analyzed for security threats as it happens. No configuration required — protection is always on.
Prompt injection detection
Catches attempts to override system instructions, extract prompts, or manipulate agent behavior.
Jailbreak attempt detection
Identifies role-play attacks, encoding tricks, and other jailbreak techniques.
RAG poisoning detection
Flags embedded instructions in retrieved documents designed to hijack agent responses.
NSFW content detection
Detects inappropriate content in both user messages and agent responses.