AgentKit Core

Plugin agentkit-core in RemseyMailjard/agentkit. Commands read from the repo's manifest — 1 month ago.

Skills (119)

• analyze-adoption

  Use when a user wants to understand why a technology, AI tool or process may not be adopted and what interventions are likely to improve usage.

• analyze-decision

  Use when a user wants to identify what was decided, what remains an assumption, and whether notes, messages or requirements contain conflicting statements.

• analyze-incident

  Use when an outage or production failure needs an evidence-based timeline, root cause, impact and preventive actions.

• analyze-requirements

  Use when a user provides a feature request, client request, brief or requirement set and wants it made implementation-ready and testable.

• assess-agent-readiness

  Use when an AI agent should be assessed for pilot or production readiness.

• assess-agentkit-release

  Use when AgentKit should be assessed for a versioned release across structure, routing, eval coverage and deprecations.

• assess-ai-product-opportunity

  Use when an AI/agent idea should be assessed for real product value, frequency, feasibility and safer non-AI alternatives.

• assess-automation-opportunity

  Use when a business process should be assessed for Power Automate, agentic automation or another workflow automation approach.

• assess-automation-scale

  Use when an automation's likely volume, cost and bottlenecks should be assessed before scaling.

• assess-mcp-release-readiness

  Use when an MCP server should be assessed for pilot or production readiness.

• assess-product-opportunity

  Use when repeated consulting, training or delivery work may be turned into a reusable product, workshop, service or subscription.

• assess-training-readiness

  Use when a training, workshop or hands-on session must be checked for real-world delivery readiness, including labs, access, tooling and fallback plans.

• audit-agentkit-routing

  Use when AgentKit Skill/plugin routing should be checked for ambiguous triggers and boundary regressions.

• audit-knowledge-quality

  Use when a knowledge base, memory set or project context should be reviewed for staleness, duplication, unsupported claims and weak provenance.

• audit-person-memory

  Use when stored person context should be reviewed for relevance, duplication, sensitivity and unsupported inference.

• build-case-study

  Use when completed client work should become a credible case study without exaggeration.

• build-mcp-test-plan

  Use when an MCP server needs a practical test matrix across protocol, backend, auth and failures.

• build-operational-runbook

  Use when a technical process needs an executable runbook with troubleshooting and escalation.

• build-person-context

  Use when provided messages or notes should be turned into a respectful, durable person-context summary.

• build-task-context

  Use when a task needs a compact relevant context package assembled from a larger body of knowledge or project history.

• build-training-module

  Use when one part of a technical training should be designed from concept through explanation, example, demo, guided practice and challenge.

• check-definition-of-done

  Use when a user wants to verify whether code, a document, workflow, project result or other artifact is genuinely complete rather than merely implemented.

• choose-daily-focus

  Use when the user wants to choose one meaningful outcome to finish today while protecting focus and energy.

• close-consulting-engagement

  Use when a consulting engagement should be closed with acceptance, value evidence, open items and reusable IP captured.

• close-work-item

  Use when a user wants to determine whether a project, task, delivery or client item is truly ready to close.

• compare-options

  Use when a user wants to compare technologies, architectures, vendors or implementation approaches and receive a concrete recommendation.

• compound-learning

  Use when new knowledge should be converted into reusable teaching, consulting or product assets rather than remain passive learning.

• compress-context

  Use when a large conversation, project history or document set must be compressed without losing decisions, constraints, evidence and next actions.

• conduct-decision-research

  Use when research should support a concrete technical, product or business decision rather than produce a generic information summary.

• curate-knowledge

  Use when notes, research or project information should be cleaned, deduplicated and stored as durable reusable knowledge rather than raw history.

• define-analytics-question

  Use when a vague reporting or analytics request should become a decision-oriented data question.

• design-agent-evals

  Use when an agent, MCP workflow or AI automation needs practical evaluation cases for routing, tools, failures and safe behavior.

• design-agent-product

  Use when an AI agent should be designed as a product with user journey, trust, metrics, feedback and governance built in.

• design-api-integration

  Use when an API-backed integration needs capability mapping, auth, contracts, resilience and deployment-aware architecture.

• design-automation-fallback

  Use when a critical automation needs safe manual recovery and reconciliation behavior.

• design-automation-operations

  Use when an automation needs monitoring, SLA, ownership and operational support design.

• design-azure-integration

  Use when an integration or backend capability needs a minimal Azure architecture with identity, security, operations and rollback.

• design-dotnet-service

  Use when a .NET service or API should be designed with clear boundaries, DI, configuration, observability and testability.

• design-enterprise-rollout

  Use when a user wants a practical rollout plan for a new AI, Microsoft or digital capability in an enterprise environment.

• design-event-driven-flow

  Use when a workflow may benefit from events, queues, asynchronous processing or decoupled integration.

• design-governance

  Use when a new AI, Microsoft or digital capability needs proportionate governance, ownership and minimum policy controls.

• design-learning-sprint

  Use when a technical topic should be learned quickly through hands-on work with clear proof of competence.

• design-m365-solution

  Use when a business process or productivity problem should be translated into a practical Microsoft 365, Power Platform or Copilot solution architecture.

• design-mcp-capabilities

  Use when an API or backend should be translated into a clean MCP capability model.

• design-mvp

  Use when a user wants to turn an idea into the smallest practical MVP that can validate value, feasibility and demand.

• design-offering

  Use when a consulting, training or AI capability needs to become a sellable package, workshop, service or subscription.

• design-pilot

  Use when a product, AI capability, automation or enterprise change should be validated through a small controlled pilot before scale-up.

• design-power-platform-solution

  Use when a Power Apps, Power Automate or Dataverse solution needs architecture, environment, governance and ALM decisions.

• design-quarter-focus

  Use when many goals should be reduced to a focused quarterly direction with a small set of compounding initiatives.

• design-trainer-demo

  Use when a technical concept needs a reliable instructor demo with clear talking points, expected observations and fallback behavior.

• diagnose-technical-issue

  Use when a technical problem, error or unreliable behavior must be narrowed to the most likely cause and next diagnostic step.

• extract-reusable-ip

  Use when completed project, training, consulting or development work should be mined for reusable assets, templates, Skills, labs or products.

• final-message-check

  Use when a user wants a fast final check of an email, chat message or other outbound text before sending, focused only on material risks and clarity.

• find-next-action

  Use when a user provides an email, note, meeting outcome, project item or conversation and wants the next concrete action or waiting-for state.

• generate-test-strategy

  Use when a user wants a practical test strategy derived from requirements, risks, changes or acceptance criteria.

• grow-client-account

  Use when existing client work should be reviewed for follow-on opportunities, relationship actions and reusable value.

• identify-training-expansion

  Use when completed training may lead to additional modules, workshops or implementation support.

• improve-observability

  Use when debugging or operations are hindered by missing logs, metrics, traces or correlation data.

• manage-commercial-pipeline

  Use when multiple leads and opportunities should be prioritized into a focused commercial pipeline.

• manage-consulting-scope

  Use when a consulting engagement needs clear scope control, assumptions and risk tracking.

• optimize-agent-economics

  Use when an agent's latency, token/tool cost or operational overhead should be balanced against task value.

• optimize-agentkit-context

  Use when AgentKit token/context cost should be reduced without losing capability quality.

• plan-change

  Use when a user wants to implement a technical or process change safely, including impact, dependencies, verification and rollback.

• plan-content-campaign

  Use when one core topic should be turned into a small coherent set of content assets tied to a business goal.

• plan-content-reuse

  Use when a source artifact such as training, research, code, notes or a client solution should be repurposed into multiple useful outputs.

• plan-training-day

  Use when learning objectives and modules should be turned into a realistic instructor-led day plan with enough practice and contingency.

• position-workshop

  Use when a workshop or training product needs a clear promise, audience, business value and realistic outcome.

• prepare-change-communication

  Use when a release, rollout or process change should be communicated clearly to users and stakeholders.

• prepare-client-handoff

  Use when a consulting solution should be handed over with ownership, documentation and capability transfer.

• prepare-client-status

  Use when project information should be turned into a concise stakeholder status update.

• prepare-continuation

  Use when work must be paused or transferred and the user wants a clean continuation checkpoint for a future session or another AI.

• prepare-executive-brief

  Use when complex technical or project information should be converted into a concise management decision brief.

• prepare-handoff

  Use when a user wants to transfer project, task or conversation context to another person or AI so work can continue without rereading the full history.

• prepare-person-followup

  Use when the user wants context and the next useful follow-up before contacting someone.

• prepare-power-platform-release

  Use when a Power Platform solution must be promoted safely across environments.

• prepare-proposal-strategy

  Use when an active opportunity needs a proposal strategy before writing the actual proposal.

• prepare-security-approval

  Use when a technical change needs a concise security approval package rather than an open-ended review.

• prepare-training-delivery

  Use when an upcoming client training needs a complete operational readiness check.

• prioritize-agent-roadmap

  Use when an AI agent roadmap should be phased from prototype through pilot to production based on value and evidence.

• prioritize-learning-radar

  Use when technology updates, courses or topics should be filtered into a focused learning radar based on real work relevance.

• prioritize-work
• process-meeting-outcomes
• qualify-opportunity
• qualify-training-request
• rebalance-personal-load
• record-architecture-decision
• review-agent-design
• review-agentkit-capabilities
• review-analytics-access
• review-api-architecture
• review-aspnet-endpoints
• review-automation-reliability
• review-dashboard-value
• review-data-quality
• review-decision-quality
• review-dotnet-implementation
• review-ef-data-access
• review-fix-plan
• review-initiative-portfolio
• review-mcp-contracts
• review-power-app
• review-power-automate-flow
• review-power-platform-solution
• review-privacy-impact
• review-security-posture
• review-training-content
• review-training-delivery
• run-strategic-review
• run-weekly-review
• secure-agent-workflow
• shape-client-proposal
• shape-linkedin-post
• summarize-relationship-context
• systemize-personal-routine
• triage-personal-inbox
• update-training-from-radar
• validate-claims
• verify-security-controls
• write-technical-guidance

Similar plugins