Run your first assessment with the CLI

Check setup, inspect the Catalog and agents, create a journey, run it, and review the evidence-backed result.

This workflow uses the short command aliases. Replace example IDs with values returned by your workspace.

Check workspace status

ivk st
ivk setup show

Status explains readiness and the next action. Setup shows product-environment and onboarding state.

Inspect product connections

ivk catalog connections
ivk catalog discovery

If no connection exists, request discovery with structured input:

ivk catalog discover \
  --field productSurfacePublicId=srf_example \
  --field baseUrl=https://api.example.com

Add a required credential

ivk catalog credential pdr_example --secret-file ./token.txt

For standard input:

printf '%s' "$PRODUCT_TOKEN" | \
  ivk catalog credential pdr_example --secret-file -

The CLI seals the credential locally and does not print the plaintext.

Inspect the Catalog and agents

ivk catalog list
ivk agents list
ivk agents ensure-managed

Review any action or agent that needs setup before continuing.

Create a journey

ivk j create \
  "A customer asks the agent to create a project and confirm the final project state"

The command may ask for clarification. Answer with:

ivk j create --answer "The customer must see the project in the workspace list"

Review and run the journey

ivk j list
ivk j get jny_example
ivk j run jny_example --wait

--wait follows assessment planning until it reaches the command's wait boundary. If the command returns before execution finishes, use the plan ID with ivk r progress.

Inspect results and findings

ivk r list
ivk r get run_example
ivk f list
ivk f get fnd_example
ivk f fix --field findingPublicId=fnd_example

Treat the verified checks as the result. The agent's final message is not proof by itself.

Use JSON in scripts

ivk --json st
ivk --json j list
ivk --json r get run_example

Place global options before the command group for predictable parsing.

See Structured input and output and Command reference.

Run your first assessment with the CLI

On this page