Agent Skills by kgan01
Showing 52 vetted skills indexed across 1 GitHub repositories.
Kgan01general-research
Evaluate sources and synthesize evidence into structured research conclusions.
Kgan01hallucination-detector
Detect fabricated URLs, unsourced statistics, fake citations, and unfounded future claims in agent outputs.
Kgan01agent-identity
Build an evolving user model from observed preferences and communication styles.
Kgan01home-lighting
Plan home lighting setups for mood, energy efficiency, and circadian alignment.
Kgan01music-curation
Generate mood-aware playlists using genre classification, BPM matching, and energy/valence constraints.
Kgan01tutoring
Guide learners through concepts using Socratic questioning and adaptive difficulty.
Kgan01task-tracking
Organize tasks and projects using GTD and Eisenhower matrix workflows.
Kgan01mcp-patterns
Unifies MCP server connections and tool routing via a meta-tool pattern.
Kgan01devops
Automates Docker and GitHub Actions deployments for solo developers and small projects with rollback-ready scripts.
Kgan01shopping
Compare product prices and reviews across multiple sources into three tiers.
Kgan01bookkeeping
Automate financial record-keeping, expense categorization, and tax preparation for small businesses.
Kgan01report-writing
Generate structured executive reports with citations from research findings.
Kgan01skill-chain-supervisor
Coordinate and validate multi-skill workflows using a shared JSON scratchpad.
Kgan01scheduling
Organize time blocks, rituals, and priority-based schedules across calendars.
Kgan01execution-harness
Orchestrate long-running projects across Claude Code sessions with checkpointing and resume capability.
Kgan01skill-memory
Organize domain knowledge into grepable markdown entries with timestamps.
Kgan01output-formatting
Format content for eight destination formats with deterministic source-to-output mapping.
Kgan01time-perception
Inject timestamps and elapsed time into LLM prompts with JSONL logging.
Kgan01constitutional-ai
Enforce constitutional safety rules across autonomous agent workflows.
Kgan01auto-project-sync
Update CLAUDE.md, MEMORY.md, and project indexes after work batches.
Kgan01deep-research
Organize iterative research into a seven-phase workflow with gap tracking and red teaming.
Kgan01context-health
Detect context degradation and apply recovery strategies in long-running tasks.
Kgan01setup
Configure Claude Code autonomous permissions with allow and deny lists.
Kgan01paper-to-code
Extract engineering plans from research papers into delta specs.