loxihart-ai

Plugin loxihart-ai from marketplace loxihart-skills in Krosebrook/loxihart-skills. Commands read from the repo's manifest — today.

Skills (37)

• ai-agents-workflow

  Builds and implements AI agents, automation workflows, and tool integrations — MCP servers (2025 async, OAuth, stateless spec), LangChain and CrewAI agents, multi-LLM orchestration, and no-code platforms. Use when the user mentions: MCP server, build an agent, tool integration, API orchestration, mu

• ai-solution-architecture

  Design pre-sales technical architectures for AI/ML solutions targeting B2B enterprise clients. Use when scoping AI implementations, designing reference architectures, mapping integration patterns, estimating infrastructure requirements, or preparing technical proposals. Triggers on solution design,

• applying-adversarial-review

  Applies Kyle's adversarial review methodology to consequential outputs. Forces every recommendation, audit, plan, or claim through three checks before shipping: a CLAIMS / COUNTEREXAMPLE / CONTRADICTIONS footer, [SOURCED] vs [GENERATED] provenance labeling, and an agreement-streak flag at three cons

• auditing-agent-security

  Applies the Agent Security framework to autonomous AI systems and ghost-team architectures — permission and blast-radius mapping, prompt-injection and context-poisoning review, MCP bridge security, and quarterly cleanses. Use when the user mentions: agent security audit, permission stack review, age

• auditing-cross-lane-routine-artifacts

  Adversarially audits multi-routine automation artifacts (cron/API routine libraries, scheduled-task decks, "routine cards") for cross-lane contamination, scheduled-task governance violations, and the structurally-empty-at-runtime defect — then splits them into lane-clean artifacts and rebuilds them

• avoiding-ai-slop

  Detects and eliminates AI-slop from text, prose, and image prompts using a 76-rule structural pattern corpus, a 237-term tiered ban lexicon, and an image-slop negative-prompt pack, with a slopcheck.py linter for scan, calibrate, and allowlist workflows. Use when the user mentions: de-slop this, desl

• briefing-loxihart-leadership

  Prepares and structures leadership communications for Kyle Rosebrook's AI SME role at Loxihart — monthly progress updates, initiative proposals, adoption milestone announcements, and domain unlock requests. Tailored to three named stakeholders: [PERSON] (direct manager), [PERSON], and [PERSON] Miles

• building-langgraph-agents

  Builds stateful AI agents using LangGraph — the graph-based state machine framework for agents with complex conditional flows, retry loops, error recovery, and built-in HITL checkpoints. Covers when to choose LangGraph over LangChain or CrewAI, core concepts (StateGraph, TypedDict state, nodes, edge

• capturing-meeting-intelligence

  Connects Otter.ai (transcription), Granola (structured meeting notes), and Mem (knowledge storage) into a unified meeting intelligence workflow for Loxihart and FlashFusion. Routes each meeting to the right capture tool, runs a post-meeting processing protocol within 24 hours, and stores distilled c

• coaching-colleagues-on-ai

  Guides informal AI coaching conversations with Loxihart colleagues — from first questions to skeptic conversations to helping power users go deeper. Use when a colleague asks about AI tools, when someone is skeptical about AI at work, when a teammate needs help getting started with Claude, when some

• correcting-ai-generated-content

  Audits AI-generated documents, NotebookLM exports, and Google Drive research files for factual gaps, missing production constraints, outdated information, and unverified claims. Runs web search to fill each gap. Rebuilds a corrected version in the target brand voice with [SOURCED]/[GENERATED] taggin

• create-meta-prompts

  Creates optimized prompts for Claude-to-Claude pipelines with research, planning, and execution stages. Use when building prompts that produce outputs for other prompts to consume, or running multi-stage workflows (research → plan → implement). Triggers on: meta-prompt, prompt chain, Claude pipeline

• create-subagents

  Expert guidance for creating, building, and using Claude Code subagents and the Task tool. Use when working with subagents, setting up agent configurations, understanding how agents work, or using the Task tool to launch specialized agents.

• deep-research-lcot

  Use for deep research, literature reviews, technical due diligence, scientific explainers, market/technical reports, and reasoning-first knowledge synthesis. Turns a question into a verified research plan, evidence matrix, derivation graph, contradiction audit, and cited final report. Activates on:

• designing-hitl-workflows

  Designs human-in-the-loop (HITL) checkpoints for AI agent workflows — deciding where to pause for human approval, structuring the approval interface, handling timeouts and non-responses, and building escalation paths from automated to human. Covers implementation patterns for n8n wait nodes, Claude

• designing-research-surveys

  Designs, deploys, and analyzes surveys for four recurring FlashFusion and Loxihart use cases: AI adoption measurement (Loxihart), post-engagement client feedback (FlashFusion), workshop and training feedback (both), and audience/content research (FlashFusion newsletter). Pulls pre-built question tem

• developing-skills

  Provides authoritative guidance for creating and managing Claude skills following Anthropic best practices across both Cowork plugin skills and Claude Code skills. Covers skill structure, YAML frontmatter, trigger phrase density, gerund naming, progressive disclosure patterns, evaluation-driven auth

• enabling-ai-solutions

  Designs and scopes AI solution architectures for personal projects, platform builds, and internal tooling. Use when mapping technical requirements to AI stack decisions, selecting LLM providers, sizing infrastructure, planning integrations, modeling costs, or producing a spec for any AI-enabled buil

• enforcing-output-formats

  Enforces industry-standard formatting conventions for every output type, even when not explicitly requested. Use for any request to produce a song, lyric sheet, script, screenplay, email, letter, memo, legal document, resume, recipe, chord chart, lead sheet, tablature, lesson plan, press release, in

• evaluating-ai-agents

  Designs and runs evaluations (evals) for custom AI agents to measure behavioral quality, catch regressions, and establish production readiness. Based on Anthropic's "Demystifying evals for AI agents" (Jan 2026). Covers eval structure (task/trial/transcript/outcome), grader selection (code-based, mod

• facilitating-ai-workshops

  Facilitates AI enablement workshops and consulting exercises that move stakeholders from abstract AI concepts to measurable business outcomes. Use when running an AI workshop, facilitating AI adoption sessions, running consulting exercises, helping teams identify AI use cases, mapping AI to business

• ff-agent-stack-v199

  FlashFusion Agent Stack v1.9.9 phased build pipeline. Activates for: agent stack, v1.9.9, phase 0, phase 1, phase 2, phase 3, phase 4, gaps audit, baton pass, langgraph, orchestration build, agent pipeline, flashfusion, flash-n-frame, supabase schema, rls policies, agent catalog, graph spec, memory

• generating-audience-tier-reports

  Use when the same underlying research or dataset needs to ship as multiple audience-specific documents from one source of truth — e.g. "write this for investors, for students, and for the general reader," "make an entrepreneur version and an internal-memo version of this report," "same data, differe

• instrumenting-agent-telemetry

  Instruments AI agents to produce observable telemetry — what to log at each agent step, which metrics signal health vs. degradation, how to wire PostHog (MCP connected) for agent-specific observability, and how to detect behavioral drift before it reaches users. Datadog is NOT connected — all Datado

• loxihart-ai-enablement-pm

  Assist with AI enablement product management at Loxihart — writing AI adoption PRDs, stakeholder briefings for [PERSON], [PERSON], and [PERSON], feature analysis for the Loxihart AI platform, roadmap prioritization using WSJF, domain unlock rollout proposals, and AI skill launch planning. Use when w

• loxihart-data-analyst

  Analyze data from Loxihart's operational stack — PostHog product analytics, Supabase database queries, n8n workflow logs, domain unlock tracking, and Stripe revenue data. Use when analyzing AI adoption metrics, token consumption, user behavior, workflow performance, or building data-backed reports f

• managing-agent-context

  Manages context windows in long-running AI agent loops — deciding when to summarize vs. truncate, implementing rolling window patterns, preserving critical context while shedding stale state, and budgeting tokens across multi-turn agent interactions. Use when an agent is running out of context, buil

• orchestrating-ai-agents

  Architecture and design guidance for AI agents, automation workflows, and tool orchestrations — topology selection, trust and permission boundaries, failure-mode and rollback design across the AI/automation ecosystem. Use when the user mentions: agent architecture, how should I structure this agent

• packaging-skills

  Packages Claude skills into production-ready zip files with full March 2026 Anthropic spec compliance, including structure validation, YAML verification, gerund naming enforcement, path reference validation (all SKILL.md subdirectory paths must exist in the zip), version sync checks, and exclusion o

• producing-ai-newsletter

  Operates a multi-configuration AI newsletter production system with shared workflow engine (research, draft, edit, format, deliver) and switchable brand profiles. Ships with Loxihart configuration (3-tier audience, Teams and SharePoint delivery, continuity tracking) and FlashFusion configuration (si

• reminding-skill-usage

  Analyzes your task and suggests which skills from your library would be most helpful. Use when starting a new task or unsure which existing skills apply to your current work.

• roi-calculator-building

  Build AI implementation ROI models for B2B enterprise sales. Use when creating ROI calculators, building business cases for AI adoption, estimating cost savings from automation, projecting revenue impact of AI features, or preparing financial justification for AI projects. Also triggers on ROI analy

• running-drive-automation

  Operates the Drive Automation Chain 1+2 system — a multi-agent Google Drive classification and routing pipeline backed by a Supabase audit trail. Activates when the user wants to classify, route, or organize Google Drive files; invoke the agent council; run the drive router; check for file drift; tr

• scoping-ai-engagements

  Scopes, proposes, and phases AI enablement engagements for FlashFusion client organizations. Use when running a discovery conversation, writing a client PRD, scoring a WSJF backlog, building a domain-unlock or capability rollout, drafting a stakeholder brief, or turning discovery answers into a phas

• task-decomposition

  Decompose complex multi-part requests into parallel workstreams with dependency tracking and integration planning. Use this skill whenever a request involves 3 or more distinct components, domains, or deliverables that could be worked on independently — even if the user doesn't explicitly ask for de

• use-case-discovery

  Runs FlashFusion AI use-case discovery workshops with enterprise clients to produce a ranked, scored backlog of automation and AI opportunities. Use when facilitating a discovery call, running an AI ideation or strategy workshop, assessing a department for automation candidates, scoring a list of us

• using-fart-prompt

  Breaks creative, writing, planning, and thinking paralysis by generating intentionally ugly, messy, half-baked versions of an idea before the polished version arrives. Lowers activation energy by making "bad" the goal. Use when stuck, blocked, overthinking, or paralyzed by perfectionism. Trigger phr

Similar plugins