AI Operating Partner

Plugin ai-operating-partner-model in kavins06/AI-OPERATING-PARTNER-MODEL. Commands read from the repo's manifest — 1 month ago.

Skills (27)

• agent-architecture-and-tooling

  Translate approved handoff artifacts into implementation-grade architecture, tool, data, and identity contracts. Use after approved Step 18 handoff and separate implementation approval as part of the Section 2 build and Managed AgentOps backbone.

• agent-build-intake-and-scope

  Gate implementation approval and lock the first safe agent behavior. Use after approved Step 18 handoff and separate implementation approval as part of the Section 2 build and Managed AgentOps backbone.

• agent-build-managed-agentops

  Run the post-18 agent build and Managed AgentOps backbone after an AI operating partner engagement has completed Step 18, the Step 17 implementation decision packet and Step 18 managed lifecycle or lifecycle_set are approved, the method-to-build handoff is accepted, and a separate implementation app

• agent-change-expansion-retirement

  Review material changes, expansion, rollback, remediation, retirement, and redesign loops. Use after approved Step 18 handoff and separate implementation approval as part of the Section 2 build and Managed AgentOps backbone.

• agent-environment-and-access

  Control environment progression, test data, secrets, provisioning, access review, and rollback readiness. Use after approved Step 18 handoff and separate implementation approval as part of the Section 2 build and Managed AgentOps backbone.

• agent-evals-guardrails-observability

  Design implementation-grade eval suites, guardrail controls, security-control evidence, trace policy, dashboards, alerting, and incident rehearsal for Section 2 phases 7 and 9. Use after approved Step 18 handoff and separate implementation approval as part of the Section 2 build and Managed AgentOps

• agent-release-management

  Version, package, approve, release, monitor, and roll back approved agent artifacts for Section 2 phase 12, including prompts, models, tools, schemas, eval sets, guardrails, trace specs, permissions, release notes, smoke tests, feature flags, deployment evidence, rollback package, and production mon

• agent-sandbox-pilot-readiness

  Govern sandbox prototype evidence, controlled pilot launch, production readiness, incident rehearsal, revocation readiness, monitoring gates, and pilot-to-production promotion for Section 2 phases 8, 10, and 11. Use after approved Step 18 handoff and separate implementation approval as part of the S

• ai-agent-readiness-assessment

  Create Step 16 implementation-grade readiness_object artifacts for candidate opportunities by evaluating hard gates, behavior-level readiness, use-case framing, business value, workflow clarity, source access, data quality, truth production, knowledge/guidance, measurement, AI-capability metrics, so

• ai-operating-partner-engagement

  Orchestrate a complete AI operating partner engagement from executive opportunity terrain mapping through AI-led interviews, source access intelligence, AI-native workflow specification, controlled validation/governance resolution, knowledge/analytics/risk enrichment, architecture/integration bluepr

• ai-operating-partner-start

  Start or resume AI Operating Partner in ChatGPT Work or Codex. Use for getting started, finding AI opportunities, choosing the right workflow, assessing a business process, or routing an approved implementation handoff. Gives nontechnical users one entry point into the bundled discovery and AgentOps

• analytics-maturity-assessment

  Assess an organization's analytics maturity and measurement intelligence across reporting, data integration, metric trust, governance, semantic definitions, predictive capability, dashboards, and agent readiness. Use when Codex needs to score current state, classify trusted measurements, identify ga

• dashboard-kpi-design

  Design decision-ready dashboards and KPI standards with audience, business question, formulas, grain, official sources, de facto trusted sources, truth production chains, context, thresholds, filters, owners, and quality checks. Use when Codex needs dashboard specs, KPI dictionaries, metric governan

• data-governance-discovery

  Discover and document Step 8 object-driven governance requirements including official sources, de facto trusted sources, truth production chains, source access paths, owners, stewards, definitions, data quality, access rights, retention, lineage, baseline release rules, lifecycle implications, and a

• data-monetization-canvas

  Create Step 12 business value cases and portfolio comparisons for data, analytics, and AI opportunities through internal efficiency, customer experience, decision quality, data products, partnerships, risk reduction, revenue, leakage prevention, or retention. Use when Codex needs an evidence-linked

• data-strategy-workshop

  Plan, run, and synthesize Step 1-2 strategy workshops that translate business goals into executive opportunity terrain, data needs, governance choices, analytics opportunities, AI-agent candidates, candidate-use-case shortlists, and measurable modernization priorities.

• information-ecology-diagnostic

  Diagnose a Step 7 business workflow or information environment across people, process, culture, incentives, politics, architecture, behavior, tools, data flows, truth production, adoption, and regulated-domain constraints before recommending analytics, automation, or AI agents.

• knowledge-map-capture

  Create the Step 9 AI knowledge and guideline requirements map by identifying decision-linked knowledge areas, likely experts, documents, tacit and explicit knowledge, regulated-domain guidance needs, gaps, risks, owners, and next capture routes before training, automation, analytics, or AI-agent des

• managed-agent-lifecycle

  Create Step 18 machine-readable managed_lifecycle_object or lifecycle_set artifacts for future AI capabilities, remediation paths, or no-automation review cadences, covering ownership, approval, launch gates, validation, monitoring, feedback, truth governance, source freshness, access review, incide

• managed-agentops-operations

  Operate live agents through metrics, source freshness, incidents, access review, guidance updates, and value tracking. Use after approved Step 18 handoff and separate implementation approval as part of the Section 2 build and Managed AgentOps backbone.

• privacy-security-risk-register

  Create the Step 15 implementation-grade risk_control_model for privacy, security, compliance, operational, vendor, output-quality, data-handling, truth-production, regulated-domain, adoption, and AI-agent risks before implementation.

• source-grounded-consulting

  Create traceable, evidence-based consulting recommendations by separating facts, assumptions, inferences, and recommendations. In the AI operating partner engagement, use as the cross-step evidence discipline for Steps 0-18, sponsor-facing views, packet/lifecycle recommendations, and current law, re

• tacit-to-explicit-knowledge-capture

  Convert Step 9 knowledge requirements and expert tacit knowledge into Step 10 AI guidance packs: structured rules, source hierarchy, truth-production handling, examples, checklists, SOP views, escalation criteria, output contracts, and behavioral/output-quality eval cases.

• voice-interview-discovery

  Plan, conduct, and synthesize Step 3 role-aware AI-led discovery interviews to understand how work actually happens before data, analytics, automation, or AI-agent design. Use when Codex needs interview guides, transcript extraction, work-episode analysis, evidence-need logs, source-access signals,

• warehouse-lakehouse-blueprint

  Design the Step 14 technical_blueprint for future analytics, dashboards, data products, and AI agents, including warehouse/lakehouse, semantic layer, truth production remediation, document intelligence, retrieval, integration, runtime/orchestration requirements, hosting/environment requirements, MCP

• workflow-mapping

  Convert interviews, notes, transcripts, SOPs, reports, or observed work into the Step 6 workflow_intelligence_object. Use when Codex needs structured organizational workflow intelligence: roles, hierarchy, actors, steps, decisions, handoffs, exceptions, information objects, source access profiles, t

• zero-trust-agent-controls

  Design the zero-trust portion of the Step 15 risk_control_model for future AI agents, automation, analytics tools, and data workflows, including least privilege, scoped credentials, identity, application, data, automation, visibility, approval gates, logging, monitoring, incident response, and revoc

Similar plugins