Integrations · Cypress
A Guild agent reads Cypress Cloud test reporting data in natural language, without handling API credentials directly. Guild stores the Cypress API key server-side and injects it at request time, so it never enters agent code or prompts. Credential policies let you restrict which reports each agent can request, and every call appears in the audit trail.
Book a demo
See how teams govern agents that use Cypress.
Start free
Connect Cypress and scope your first agent.
What the Cypress integration exposes
The integration exposes a single tool,cypress_call, which accepts a natural language instruction and dispatches it to the appropriate Cypress Cloud report. Available reports cover project lists, spec details, flake rate over time, usage per project, and test suite size. All reports are read-only. No tool creates or manages test runs.
Get started
1
Open Credentials in Guild
Go to Credentials in Guild under Access & setup. This is where you connect and scope all third-party service credentials.
2
Enter your Cypress API key
Click Cypress and enter your Cypress Cloud API key. Find the key in Cypress Cloud under Organization Settings > Enterprise Reporting. Enterprise Reporting API access is required and is not available on all Cypress plans.
3
Import and spread cypressTools
Import
cypressTools from @guildai-services/guildai~cypress and spread it into your agent’s tools object. The agent can then request Cypress Cloud reports in natural language.Why Cypress Cloud access needs governing
Cypress API keys are long-lived and carry no built-in expiry. A credential scoped at the account level serves every workspace the account owns by default. Guild lets you narrow that scope to a single workspace, a single member, or one specific agent, so the key reaches only the agents that need it. See Credentials for the full scoping model. Credential policies add a second layer of control. Guild’s credential proxy evaluates ALLOW and DENY rules before each request leaves the platform. A DENY rule blocks a request even when an ALLOW rule also matches, and the agent cannot bypass policy because the credential is injected server-side only after the policy check passes. See Credential policies for rule syntax. Without scoped policies, a new credential starts with an allow-all default.About the Cypress integration
The@guildai-services/guildai~cypress package provides agents with read-only access to Cypress Cloud test reporting data through an LLM-mediated interface.
Authentication
- Type: API key
- Token expiration: None — API keys are long-lived
Setup
1
Open Credentials
Go to Credentials in Guild.
2
Connect Cypress
Click Cypress and enter your Cypress Cloud API key.
3
Locate your key
Find your API key in Cypress Cloud under Organization Settings > Enterprise Reporting.
Enterprise Reporting API access is required. This is not available on all Cypress plans.
Usage
How it works
Unlike other services that map directly to REST endpoints, the Cypress service uses an LLM-mediated approach. Agents describe what data they need in natural language, and the system selects the appropriate report.Tools
Tools in the latest published version ofguildai~cypress:
Available reports
Thecypress_call endpoint dispatches to these reports based on the instruction:
All date parameters use
yyyy-MM-dd format.
Webhook events
Cypress does not support webhook triggers in Guild. This integration is API-access only.Limitations
- Read-only reporting data — no creating or managing test runs
- Request timeout is 60 seconds per report
- No webhook or trigger support
Frequently asked questions
Can I trigger an agent run automatically when a Cypress test suite finishes?
Can I trigger an agent run automatically when a Cypress test suite finishes?
No. Cypress does not support webhook triggers in Guild. The integration is API access only, so agents must be invoked directly rather than by an event from Cypress Cloud.
What date range can I pass to a Cypress report?
What date range can I pass to a Cypress report?
Each report requires a
start_date parameter and accepts an optional end_date. Both use the yyyy-MM-dd format. The cypress_call tool selects the correct report and parameters from your natural language instruction.Does the agent ever see my Cypress API key?
Does the agent ever see my Cypress API key?
No. Guild stores the key and adds it to outbound requests server-side through its credential proxy. The key never enters agent code, the agent container, or the prompt.
Can I limit one agent to only flake rate data and block another from it?
Can I limit one agent to only flake rate data and block another from it?
Yes. Credential policies support ALLOW and DENY rules scoped to specific agents. You can write a DENY rule for one agent while an ALLOW rule covers another, all on the same connected Cypress credential. See Credential policies for rule details.
Put a governed agent on Cypress Cloud this week
Book a demo
See how teams govern agents that use Cypress.
Start free
Connect Cypress and scope your first agent.