بنقرة واحدة
sitecoreai-cli
يحتوي sitecoreai-cli على 22 من skills المجمعة من Sitecore-Studio-Labs، مع تغطية مهنية على مستوى المستودع وصفحات skill داخل الموقع.
Skills في هذا المستودع
Author Sitecore recipes (.recipe.ts) that compile to templates + renderings — map a React/SXA component to a component-template, model data with content-templates, define enums, group with component-sections. Use when creating or editing *.recipe.ts files, turning a component into a Sitecore rendering, or asking "how do I make a recipe / why doesn't my component show up in Pages".
Non-negotiable codebase conventions for the scai CLI — module boundaries, error contract, credential handling, agent contract, quality gates. Trigger when writing or modifying code in scai.
Project-specific testing patterns for scai — Vitest tiers, mocking conventions, integration gating, and where tests live. Trigger when writing tests, debugging test failures, or asking about how to test something.
Navigate scai's access model — the workspace policy allowlist and the credential/keychain patterns. Use when a command fails with POLICY_DENIED or AUTH_REQUIRED, when authenticating an environment, or to pick the fastest route to gain access to an environment.
Reference card for working on the scai CLI locally — boot sequence, common commands, env setup, sandbox tenant, and gotchas. Read this at the start of a session that involves running, building, or testing scai.
Run the CLI safely in non-interactive or CI environments. Use when the user mentions agents, CI, non-TTY, automation, or headless runs.
Guide the release flow using changesets and npm provenance. Use when publishing, versioning, or preparing a release.
Golden principles for code design — research-first culture, prefer libraries, validate at boundaries, right abstraction level. Trigger on any non-trivial implementation task.
Capture a friction moment — when the operator is frustrated, repeatedly correcting the agent, or the agent is confidently wrong. Invoke with /friction to turn frustration into a harness improvement. This skill is user-invocable.
Audit harness health — memory staleness, skill drift, and convention coverage gaps. Trigger when the user asks to "review the harness", "audit memory", "check skill freshness", or "harness health".
Bootstrap and validate sitecoreai.cli.json configuration. Use when creating new configs, adding environments, or fixing config validation errors.
Run the CLI smoke test in CI. Use when setting up or debugging the headless smoke workflow.
Run Deploy API commands for projects, environments, and deployments. Use when working with deploy workflows or managing deploy resources.
Configure authentication for environments (deploy token or client credentials). Use when the user asks about login, deploy tokens, or environment auth setup.
Manage logging, redaction, and output formats safely. Use when the user asks about logs, secrets, or JSON output.
Authors and validates serialization module configuration, includes/rules/aliases, and serialization options. Use when editing sitecoreai.cli.json or module.module.json files, or when the user asks about scopes, includes/excludes, aliases, or maxRelativePathLength.
Resolves serialization conflicts and move/rename/update issues safely. Use when pull/push conflicts, diff disagreements, or standard values ordering issues appear.
Bootstraps new serialization modules and migrations from existing items. Use when creating new module.module.json files, initial pulls, or migrating serialization layouts.
Run serialization pull/push/diff/validate/watch workflows. Use when working with serialization commands, includes/excludes, or what-if runs.
Optimizes serialization performance using include/exclude, field filters, and watch options. Use when serialization is slow, large modules are involved, or the user asks about performance tuning.
Diagnoses serialization errors and validation failures (duplicate IDs, scope/rule errors, alias conflicts, path collisions). Use when serialization validate/pull/push/watch fails or logs errors.
Diagnose common CLI failures and recommend fixes. Use when commands fail, errors are reported, or output is unexpected.