Agent-Skills.md

Agent Skills: Run and Monitor Jobs

Run and monitor jobs on B2C Commerce instances using the b2c CLI, including site archive import/export and search indexing. Use this skill whenever the user needs to trigger a job, import a site archive, export site data, rebuild search indexes, check job status, or troubleshoot failed job executions — even if they just say "import this folder" or "rebuild the search index".

UncategorizedID: salesforcecommercecloud/b2c-developer-tooling/b2c-job

Repository

salesforcecommercecloud/b2c-developer-tooling
SalesforceCommerceCloudLicense: Apache-2.0
4612

Install this agent skill to your local

pnpm dlx add-skill https://github.com/SalesforceCommerceCloud/b2c-developer-tooling/tree/HEAD/skills/b2c-cli/skills/b2c-job

Skill Files

Browse the full folder contents for b2c-job.

Download Skill

Loading file tree…

skills/b2c-cli/skills/b2c-job/SKILL.md

Skill Metadata

Name
b2c-job
Description
Run existing B2C Commerce jobs, monitor executions, inspect failures, rebuild search indexes, and import or export site archives with the CLI.

Run and Monitor Jobs

Use b2c job for existing jobs. To create job steps or author jobs.xml, use b2c:b2c-custom-job-steps. If b2c is unavailable, use npx @salesforce/b2c-cli.

Configuration and tool choice

CLI and MCP share configuration: SFCC_* variables (including project .env), project/shared dw.json, and plugins; package.json provides nonsecret defaults. Use --project-directory when running outside the target project. Inspect with b2c setup inspect --json or MCP config_inspect; both mask secrets by default. Override configured values only when needed. For resolution issues, use b2c-cli:b2c-config.

No dedicated MCP job tool exists. Prefer these CLI workflows for running/waiting on jobs and transferring archives. MCP scapi_search/scapi_execute can compose supported Admin requests; they do not replace the CLI's archive transfer workflow.

Run and inspect

b2c job run my-job --wait --timeout 600
b2c job run my-job --wait -P key=value
b2c job search --job-id my-job --count 5
b2c job wait my-job execution-id --timeout 600
b2c job log my-job --failed

job run returns immediately unless --wait is supplied. For raw system-job bodies, search indexing, standard step choices, execution deletion, and backend configuration: run and monitor. Deletion removes completed execution records; it does not cancel jobs.

Site archives

b2c job import ./archive --show-log
b2c job import ./storefront-export.zip --wait-for-storefront
b2c job export --site MySite --site-data campaigns_and_promotions --output ./export
b2c job export --storefront my-storefront --output ./export

Import/export wait by default and transfer files over WebDAV. Jobs prefer SCAPI (sfcc.jobs.rw for writes, sfcc.jobs for reads); --api-backend auto falls back on supported SCAPI rejections. See the backend section in the run/monitor reference if access fails. Check job outcome before retrying a write.

Read only the relevant reference:

  • Import remote archives or directory subsets: import options.
  • Choose export data units or output paths: export.
  • Author/validate XML and verify imported settings: b2c-cli:b2c-site-import-export.
  • Apply ordered, tracked migrations: b2c-cli:b2c-import-set-migrations.

Output

Normal CLI output suffices for status. For scripting, save large --json results to a file and print selected fields; keep the exit code and inspect failures. See export, section compact-results. Avoid dumping full execution records or reading both JSON and human output for the same result.