Manipulation gets a polite no.
Card numbers and IBANs stay out of the model.
Guardrails check each message before your agent runs and each answer before it goes out. Use the checks that ship with connic, or write your own rules in code.
What a guardrail changes in a real conversation.
The customer gets a normal reply. The run records which rule stepped in.
A customer types a card number into a carrier's app chat. The rule swaps it for a placeholder before the agent reads it, and the top-up still goes through with the saved card.
A buyer asks about a neighbour's order. A rule the team wrote checks the answer against the buyer's account and replaces it before it goes out.
A seller asks the listing assistant to write fake reviews. It declines with a fixed reply and takes no action.
Nine checks for the usual risks ship with connic.
Turn a check on and choose whether it blocks, warns or redacts. Most checks match patterns and add no model call.
Attempts to override the agent's instructions
Before the agent runsblockwarnPersonal data in requests
Before the agent runsblockwarnredactRequests outside the agent's topics
Before the agent runsblockwarnHarmful content
Before the agent and before the answerblockwarnPatterns you write yourself
Before the agent and before the answerblockwarnPersonal data in answers
Before the answer goes outblockwarnredactAnswers that leak the agent's instructions
Before the answer goes outblockwarnAnswers that miss the question
Before the answer goes outblockwarnData smuggled out in links and images
Before the answer goes outblockwarn
See every message a rule stopped.
Each check shows up in the run's record with its rule and result. Filter the runs by outcome, and a security review starts with the full list.
- In the audit log
- Every block is also an audit log entry. On enterprise, it can go on to your own monitoring tools.
- Tested before each deploy
- Your tests run the real checks, so a deploy can stop when an injection attempt gets through.
Rules sit in the agent file, next to its tools.
Input rules run before the agent, output rules after it, in the order you list them. The first rule that blocks ends the check.
version: "1.0"
name: buyer-help
description: Answers buyers' questions about their orders
model: connic/gpt-5.6-terra
system_prompt: |
Help buyers with their orders, deliveries and returns.
tools:
- orders.get_order
- orders.track_parcel
guardrails:
input:
- type: prompt_injection
mode: block
config:
rejection_message: >-
I can help with your orders, deliveries and returns.
- type: pii
mode: redact
config:
entities: [email, phone, iban, credit_card]
output:
- type: data_exfiltration
mode: block
config:
allowed_domains: [market.example.com]
- type: custom
name: order-scope
mode: block
config:
rejection_message: >-
I can only share details about orders on your account.type: prompt_injectionBlock with your own words
The customer gets the rule's
rejection_message, and the agent never sees the request.type: piiRedact before the model
Values become placeholders such as
[IBAN_REDACTED]. The agent can't use the original either.type: data_exfiltrationNo data smuggled out in links
Blocks links and markdown images that point outside
allowed_domains.type: customYour own rule in code
order-scopeis a file inguardrails/with acheck(content, context)function. It compares the answer with the buyer's order numbers from middleware. If it raises, the rule blocks.