9,081 件 ・ 関連度順
概要と使いどころ
Agent Platform Model Tuning. Use when you need to fine-tune open models or Gemini models using Agent Platform infrastructure. Don't use for model training outside Agent Platform, model deployment to endpoints (use `agent-platform-deploy`), or managing serving endpoints (use `agent-platform-endpoint-management`).
日本語の概要は準備中です。原文の説明を表示しています。
Measures and improves the quality of AI models and agents on Google Cloud using the Eval Quality Flywheel methodology. Use when evaluating an agent or model, building an eval dataset, picking or writing evaluation metrics, analyzing failures, comparing results before and after a fix, or when guidance is needed on Agent Platform eval methodology — including dataset schema, LLM-as-judge scoring, and common failure causes. For fine-tuning, use agent-platform-tuning. For general production deployment, use agent-platform-deploy.
日本語の概要は準備中です。原文の説明を表示しています。
Leverages BigQuery's built-in machine learning and GenAI capabilities for advanced data analytics. Use when you need to write SQL queries that perform time-series forecasting, predict values, detect outliers or anomalies, find key drivers, perform semantic search or vector search, classify text, calculate similarity, summarize content, translate language, evaluate models, filter by semantic conditions, measure the causal effect of an intervention, compute correlations between columns, detect change points or structural breaks, extract trend or seasonality components, or leverage generative AI capabilities in BigQuery. Do not use for general BigQuery dataset, table, or job management requests.
日本語の概要は準備中です。原文の説明を表示しています。
Review Caveman Cloud evidence read-only: costs, Cave Score, Cave Plan, workflows, traces, latency, errors, compression, routing, and verified savings. Use when the user asks what Caveman found, where LLM spend goes, why cost or quality changed, which workflows need attention, or asks for a trace or analytics review. Prefer Caveman MCP tools; fall back to CLI JSON.
日本語の概要は準備中です。原文の説明を表示しています。
Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.
日本語の概要は準備中です。原文の説明を表示しています。
Find one insight that eliminates multiple components - "if this is true, we don't need X, Y, or Z"
日本語の概要は準備中です。原文の説明を表示しています。
Use when Agentforce run costs are climbing, you need to forecast scale, or you want to reduce tokens per conversation without hurting quality. Covers topic (now subagent) design impact on cost, prompt/template reuse, grounding size discipline, caching, and model-tier selection. Triggers: 'agentforce cost', 'tokens per conversation too high', 'reduce agentforce runs spend', 'forecast agentforce scale cost', 'einstein trust layer tokens'. NOT for capping spend per user with a budget gate and fallback — use agentforce/agent-rate-limit-strategy. NOT for org-wide model-tier and BYOLLM platform strategy — use architect/ai-platform-architecture.
日本語の概要は準備中です。原文の説明を表示しています。
Author and run offline evals for Agentforce agents: fixture format, scoring rubrics, hallucination checks, regression baselines, CI integration, prompt-change safety. Trigger keywords: agent eval, hallucination, doesn't hallucinate, agentforce offline evals, rubric for agent response. NOT for deciding which layers of testing the agent needs — use agentforce/agentforce-testing-strategy. NOT for Testing Center, AiEvaluationDefinition or sf agent test runs — use agentforce/agent-testing-and-evaluation.
日本語の概要は準備中です。原文の説明を表示しています。
Use this skill when gathering, structuring, and documenting requirements for a Financial Services Cloud (FSC) wealth management implementation — including financial planning workflow discovery, portfolio review process mapping, client lifecycle requirements, advisor tooling needs, and FSC architecture determination (managed package vs. FSC Core). Trigger keywords: wealth management requirements, FSC requirements gathering, financial planning workflow, portfolio review process, advisor tools setup, FSC data model scoping, wealth management process mapping. NOT for implementation, configuration, or code — use financial-account-setup, fsc-action-plans, or apex/fsc-financial-calculations for those. NOT for FSC architecture decisions — use architect/wealth-management-architecture. NOT for Health Cloud or NPSP requirements.
日本語の概要は準備中です。原文の説明を表示しています。
Use when practitioners need to understand system-managed fields (CreatedDate, LastModifiedDate, SystemModstamp, CreatedById, LastModifiedById, IsDeleted) — their update behavior, indexing, and queryability. Triggers: 'difference between SystemModstamp and LastModifiedDate', 'how to query deleted records', 'set CreatedDate during data migration', 'why does LastModifiedDate not match SystemModstamp', 'Create Audit Fields permission'. NOT for a field's value-change history — use data/field-history-tracking. NOT for Shield Field Audit Trail — use security/field-audit-trail.
日本語の概要は準備中です。原文の説明を表示しています。
Use this skill when building, reviewing, or refreshing a RACI (Responsible / Accountable / Consulted / Informed) matrix for a Salesforce project so that every Salesforce-specific decision — data model change, automation tier choice, security model, integration boundary, deployment, license/edition — has exactly one accountable owner and a documented escalation path that downstream agents can route to. Trigger keywords: RACI matrix salesforce project, stakeholder authority salesforce, escalation path salesforce decisions, who approves data model change, salesforce decision rights, REFUSAL_NEEDS_HUMAN_REVIEW routing. NOT for the change advisory board operating model itself (use admin/change-management-and-deployment). NOT for end-user training rollout plans (use admin/change-management-and-training). NOT for the technical mechanics of permission set assignment authority (use admin/permission-set-architecture). NOT for pure stakeholder requirements elicitation (use admin/requirements-gathering-for-sf). More trigger keywords: who approves a sandbox refresh, who signs off a data load, release update activation owner, delegated admin authority vs RACI, multi-org decision rights, escalation time-box.
日本語の概要は準備中です。原文の説明を表示しています。
Use when designing or reviewing Salesforce sandbox topology, refresh cadence, masking, and environment purpose. Triggers: 'Developer sandbox', 'Developer Pro', 'Partial Copy', 'Full sandbox', 'sandbox refresh', 'data masking', 'test environment', 'Gov Cloud sandbox'. NOT for how many environments a program needs and how they map to branches - use devops/environment-strategy. NOT for re-seeding reference data after a refresh - use data/sandbox-refresh-data-strategies. Also covers: SandboxPostCopy Apex class, SandboxInfo and SandboxProcess Tooling API, SandboxSettings metadata, sandbox topology table, refresh calendar.
日本語の概要は準備中です。原文の説明を表示しています。
Use when gathering requirements for a customer portal, partner community, or self-service Experience Cloud site. Triggers: 'gathering requirements for customer portal', 'planning Experience Cloud site', 'what license for community portal', 'portal user journey mapping', 'self-service requirements'. NOT for building the site in Experience Builder — use admin/experience-cloud-site-setup. NOT for designing the deal-registration and MDF processes themselves — use admin/partner-community-requirements. More trigger keywords: portal persona licence matrix, which Experience Cloud licence do these requirements imply, external user needs a Contact and Account, sharing set vs sharing rule for a portal, self-registration vs SSO for a portal, portal requirements catalogue, handoff to Experience Cloud site setup.
日本語の概要は準備中です。原文の説明を表示しています。
Use when an admin or partner needs to package an Experience Cloud (Community) site as a reusable Lightning Bolt Solution for distribution — covers the export workflow from Experience Builder, what gets bundled (ExperienceBundle, custom apps, flow categories, theme, layouts, navigation menus) versus what does NOT (data, CMS content, files), choosing Bolt vs managed package vs unlocked package vs cloning a site, sandbox-to-production promotion, multi-org distribution, AppExchange listing as a Bolt, and template versioning via the LightningBolt metadata `versionNumber`. Triggers: 'turn this community into a reusable template', 'package an Experience Cloud site to ship to multiple orgs', 'export Experience Builder template for AppExchange', 'should we use a Bolt or a managed package for this community', 'create an industry-specific community starter', 'how do we version our partner portal template', 'distribute branded Experience site across business units'. NOT for general Experience Cloud site build, content, or member setup (use admin/experience-cloud-site-setup, admin/experience-cloud-cms-content, admin/experience-cloud-member-management). NOT for shipping Apex / LWC / data-model functionality as a product (use devops/managed-package-development, devops/second-generation-managed-packages, devops/unlocked-package-development). NOT for moving a single Experience site between sandbox and prod as a one-off (use devops/experience-cloud-deployment-admin, devops/cicd-for-experience-cloud).
日本語の概要は準備中です。原文の説明を表示しています。
Use when a BA or admin needs to document the Salesforce data model: creating field inventories, object relationship maps, ER diagrams, or analyzing field usage across objects. Triggers: 'data dictionary', 'document our data model', 'object relationship map', 'field inventory', 'ER diagram for Salesforce'. NOT for designing the model or choosing lookup vs master-detail — use data/data-model-design-patterns. NOT for generating the Mermaid or PlantUML diagram — use architect/salesforce-erd-and-diagramming. More trigger keywords: field inventory CSV, describe API, getGlobalDescribe, DescribeFieldResult, sObject describe, schema snapshot, schema drift, blank field descriptions, undocumented fields, field ownership, data dictionary review cadence.
日本語の概要は準備中です。原文の説明を表示しています。
Use when admins are governing the Chatter group lifecycle — creation policy, public/private/unlisted visibility decisions, group ownership transfer when owners leave, archived vs inactive vs deleted state changes, auto-archive behavior, naming conventions, group templates (Information Templates), and orphan-group cleanup after user terminations. Triggers: 'we have 800 dead chatter groups', 'group owner left the company who owns it now', 'should this group be public private or unlisted', 'when does a chatter group auto-archive', 'how do I delete vs archive a chatter group', 'we need a chatter group naming convention'. NOT for Chatter notification / digest / feed-noise tuning (use admin/chatter-notification-tuning), NOT for Chatter REST API integration patterns (use apex/apex-connect-api-chatter), NOT for Custom Notifications API (use apex/apex-custom-notifications-from-apex), NOT for Experience Cloud / Customer Community group access (different sharing model).
日本語の概要は準備中です。原文の説明を表示しています。
Use this skill to elicit, document, and validate CRM Analytics requirements — covering data source mapping (Salesforce object sync vs external connector vs Data Cloud), transformation needs, audience-specific lens or dashboard views, and drill-down path specifications — before any dataset or dashboard is built. Trigger keywords: CRM Analytics requirements, analytics data source mapping, CRM Analytics audience requirements, analytics visualization requirements. NOT for the formula and target behind each KPI — use admin/analytics-kpi-definition. NOT for building the dashboard once requirements are agreed — use admin/analytics-dashboard-design.
日本語の概要は準備中です。原文の説明を表示しています。
Diagnoses and fixes CI failures and re-runs GitHub Actions jobs. Use when CI is failing, tests break in CI, lint errors block merging, a build fails, or jobs need re-running.
日本語の概要は準備中です。原文の説明を表示しています。
Persistent memory patterns for moflo agents — session memory, long-term knowledge, pattern learning, and cross-session context via moflo's node:sqlite + HNSW vector store. Use when building stateful agents or assistants that need to remember across runs.
日本語の概要は準備中です。原文の説明を表示しています。
Quicken the codebase — an ad-hoc performance audit that reveals what drags (N+1 queries, needless re-renders, missing caching, memory leaks, redundant computation, hot-path complexity) and prescribes the fix, surfaced inline in the session. Scoped to your current diff by default, or to a path/area you name. Use before a perf-sensitive merge, when something feels slow, or when you want a targeted optimization pass. Replaces the old always-on `optimize` daemon worker (alias: /perf-audit).
日本語の概要は準備中です。原文の説明を表示しています。
Structured multi-hop web research with explicit confidence gating — plan the inquiry, search (WebSearch/WebFetch), score your own confidence, and keep digging until the answer is well-supported or a hop cap is hit, then emit a cited synthesis. Learns across sessions by storing each research case to memory and reusing prior strategies. Use when a question needs more than one search — comparisons, current-best-practice questions, anything where a single lookup leaves you unsure.
日本語の概要は準備中です。原文の説明を表示しています。
Detects permission requests triggered at app launch or before meaningful user interaction, which violates Guideline 5.1.1(ii) requiring permissions to be requested only at the moment of need with contextual explanation.
日本語の概要は準備中です。原文の説明を表示しています。
Record a finding about the configured engine MCP server into .unikit/MCP-RECHECK-NOTES.md — the project's log of what has to be re-checked here. Use right after a call reported success while changing nothing, ate an argument, or validated a broken state, e.g. "write this down", "record this MCP finding", "the server lied — note it", "add a recheck note", "trap this". Takes three forms of input: the finding in one line, a path to a plan file (harvests the "## MCP Findings" table of that plan and nothing else), or nothing at all — then it takes findings already in the session first, and offers to scan the tables of plans touched since the last audit. The table only, never the plan body. Writes in the one allowed genre — a check to perform; never a lifted gate, never a named workaround, never a claim about what the server can or cannot do. Makes zero MCP calls and needs no editor. To replay, retire, or upstream existing notes use /unikit-mcp-audit instead.
日本語の概要は準備中です。原文の説明を表示しています。
Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup
日本語の概要は準備中です。原文の説明を表示しています。