发现 Skills

Data structures and algorithms for AI agent episodic memory. Covers vector stores (HNSW, IVF, PQ), temporal indexing, knowledge graphs with triple stores, hierarchical summarization, forgetting curves, working/long-term/procedural memory, and memory consolidation. Deep analysis of MemGPT/Letta, Zep/Graphiti, Mem0, and the Stanford generative agents memory architecture. Teaches the CS fundamentals behind how agents remember, retrieve, and forget. Activate on: "agent memory", "episodic memory", "vector search algorithm", "HNSW", "memory retrieval", "forgetting curve", "knowledge graph memory", "MemGPT", "Letta", "Zep", "Mem0", "memory consolidation", "temporal retrieval", "agent long-term memory", "memory layer". NOT for: conversation protocol design (use agent-conversation-protocols), agent infrastructure selection (use agentic-infrastructure-2026), building RAG pipelines (use ai-engineer).

Structured interview and observation techniques for extracting tacit knowledge from domain experts

Master of high-performance web map implementations handling 5,000-100,000+ geographic data points. Specializes in Leaflet.js optimization, Supercluster algorithms, viewport-based loading, canvas rendering, and progressive disclosure UX patterns.

Techniques for surfacing implicit, hard-to-articulate knowledge that experts use in practice

Coaches behavioral and values-fit interview preparation with negative framing, deep follow-ups, introspection, and mission alignment. Use for culture-fit rounds, Anthropic behavioral prep, failure stories, and self-awareness drilling. Activate on 'behavioral interview', 'values interview', 'culture fit', 'tell me about a failure'. NOT for coding interviews, system design, resume writing, or technical deep dives.

Generate unit tests based on existing code patterns and testing frameworks.

Expert .NET 10 Full-Stack Developer skill. Use this when implementing new features, vertical slices, or modifying existing logic in the MoreSpeakers application. It covers Domain, Data (EF Core), Managers, and Web (Razor Pages + HTMX).

情感支持与心理慰藉专家，擅长倾听、共情并提供温暖的心理引导。

Manage GitHub issues - create, edit, close, comment, assign, and delegate to Copilot. Uses GitHub MCP.

Expertise in designing proactive auto-triggering components. Use when creating skills, agents, hooks, or commands that should activate automatically based on context, events, or checkpoints.

Self-evolving skills. Records user feedback and updates methodologies and rules. Trigger phrases: “update rules”, “record feedback”, “improve skill”. When user feedback or corrective trigger phrases (e.g., “remember this”, “update rules”, “that just failed”) are detected, it automatically backtracks through context to locate the issue, reads and understands target files (e.g., CLAUDE.md, tips/*.md, ../README.md), integrates corrections or new rules into the appropriate sections instead of simply appending them to the end, and keeps only event entries in the feedback log. Core features include automatic triggering, context backtracking and issue identification, full-text parsing with structured insertion, synchronized updates to routing tables and the Skill registry, and update reports with change summaries. Typical use cases include updating rules for slips of the tongue, supplementing review workflows, and iterating methodologies. Main benefits are preventing rule fragmentation, accelerating closed-loop learning, and maintaining document consistency and traceability.

Initialize gepa-research for the current repository by exploring the codebase, proposing unexplored optimization dimensions, constructing the benchmark inside a baseline worktree, and running the first experiment. Use when the user invokes /gepa-research:discover, mentions setting up gepa-research, wants to instrument a codebase for autonomous optimization, or asks to start a new gepa-research run on a project.

Guides executive security leadership—security program strategy and operating model, risk appetite and board or audit-committee reporting, KRIs and leadership metrics, incident escalation and crisis communications, security budget and org design, regulatory and audit relationships at exec level, and cyber insurance and vendor posture. Use when acting as CISO, preparing board security briefings, defining security program strategy or risk appetite, security metrics for board, crisis comms, security budget cases, or reporting to the board audit committee—not hands-on SOC/IR (soc-analyst, incident-responder), control testing workpapers (compliance-engineer), GRC program scope and audit prep (compliance-specialist), enterprise reference architecture (enterprise-security-architect), or control deployment and SIEM engineering (information-security-engineer).

Guides cybersecurity deception operations using MITRE D3FEND—honeynets, decoy objects, decoy personas, and decoy credentials. Covers honeypot deployment, decoy file planting, credential baiting, and deception environment design. Use when deploying honeypots, planting decoy data, baiting credentials, or designing deception programs—not for detection (d3fend-detect), hardening (d3fend-harden), or isolation (d3fend-isolate).

Guides cybersecurity restoration using MITRE D3FEND—recovering access, objects, configurations, and systems after incidents. Covers credential reissuance, account unlocking, file restoration, database recovery, configuration rebuild, and software reinstallation. Use after incident containment for business continuity, disaster recovery, and return to normal operations—not for incident containment (d3fend-evict), detection (d3fend-detect), or hardening (d3fend-harden).

Guides data center compute supply and resource efficiency—capacity and utilization planning, kW-per-useful-compute metrics, stranded power and rack space, hardware refresh and decommission, power-aware placement, consolidation and right-sizing of physical hosts, GPU/CPU supply alignment with workload demand, and sustainability reporting (PUE, carbon intensity). Use when optimizing on-prem or colo compute footprint, forecasting server/GPU supply, reducing idle capacity, improving DC utilization, or balancing efficiency vs performance SLAs—not for greenfield facility design (data-center-design-execution-lead), K8s deploy/troubleshoot (cluster-deployment-engineer), cloud-only FinOps (infrastructure-engineer scope varies), or LLM token cost programs (ai-token-improvement-plan-engineer), or multi-site investment prioritization and steering (data-center-portfolio-planning-execution-lead), or MW/rack construction delivery (senior-data-center-capacity-delivery-manager). Compute accounting: compute-accounting-manager.

Guides event-driven systems—pub/sub, point-to-point, event sourcing, CQRS, schema versioning (Avro/JSON Schema), ordering, partitioning, delivery guarantees, idempotency, exactly-once tradeoffs, transactional outbox/inbox, choreography vs orchestration, sagas, dead-letter queues, replay, stream vs discrete events, microservice boundaries. Use when the user asks for event-driven architecture, event sourcing, CQRS, pub/sub, Kafka events, outbox pattern, idempotent consumer, event choreography, saga orchestration, dead letter queue, or event schema versioning—not microservice code only (microservices-developer), enterprise API/iPaaS only (enterprise-integration-api-developer), EDI (edi-engineer), classified promotion (classified-software-devsecops-engineer), solvers (operations-research-algorithm-developer), generic CRUD (senior-software-engineer).

Guides FinOps analysis on AWS, GCP, and Azure—cost visibility and allocation, tagging and showback/chargeback models, rightsizing and waste removal, RI/Savings Plan/CUD recommendations, budgets and forecasts, anomaly detection, unit economics (cost per service/customer), and FinOps cadence with engineering accountability. Use when optimizing cloud spend, analyzing CUR/billing exports, building cost dashboards, explaining bill spikes, or improving allocation—not for GL mapping, capex, depreciation, or month-end ledger close (compute-accounting-manager), enterprise EA negotiation (enterprise-cloud-architect), hands-on resource provisioning (cloud-engineer), or hardware supply efficiency (data-center-compute-supply-efficiency).

Guides security assessment of embedded and cyber-physical systems on hardware-in-the-loop (HIL) test benches—bench setup, ECU/ECM or PLC targets, bus interfaces (CAN/CAN-FD, LIN, automotive Ethernet, Modbus at high level), fault injection and stimulus design, simulated plant/environment integration, attack-surface monitoring on real hardware, reproducible test cases, lab safety interlocks, and evidence capture for firmware and vehicle security teams. Use for HIL security testing, ECU security assessment, CAN bus security, PLC HIL test, fault injection lab, embedded hardware security—not web/API pentest (web-pentester), network-only pentest (network-pentester), malware/binary RE only (reverse-engineer), SOC operations (soc-analyst), AI red team (ai-redteam), classified ISSO paperwork (information-systems-security-officer-classified-specialist), or pure software CI without hardware (build-validator).

This skill should be used when the user asks about IoT network design, edge engineer work, MQTT broker and topic design, LoRaWAN, LPWAN, IoT gateway, device provisioning, IoT edge, CoAP, Azure IoT Edge, AWS IoT Core, device shadow, OTA updates, Zigbee, protocol bridge, IoT fleet, and IoT segmentation. Guides connectivity, edge gateways, store-and-forward, broker scale, and cloud IoT handoffs—not enterprise WLAN (wireless-wifi-mobility-specialist), OT/ICS plant security (scada-ics-cyber-security-specialist), embedded without connectivity (embedded-real-time-software-engineer), carrier backbone (network-backbone-architect), cloud landing zones (cloud-architect), or telemetry ML (data-scientist).

Guides microservice design and delivery—bounded contexts, service boundaries, REST/gRPC/event APIs, sync vs async tradeoffs, resilience (timeouts, retries, circuit breakers, bulkheads), per-service data ownership, saga and outbox patterns, twelve-factor containers, observability (logs, metrics, trace propagation), API versioning at gateways/meshes, and contract testing. Use for microservices developer, service boundary, bounded context, gRPC between services, circuit breaker, saga pattern, outbox pattern, twelve-factor, contract testing microservices, service decomposition, or event-driven microservice—not K8s platform ops (platform-engineer, site-reliability-engineer), enterprise iPaaS (enterprise-integration-api-developer), monolith-first apps (senior-software-engineer), or classified pipelines (classified-software-devsecops-engineer).

Expert in data pipelines, ETL processes, and data infrastructure

Create custom directives in Angular v20+ for DOM manipulation and behavior extension. Use attribute directives to modify element behavior or appearance, structural directives for portal/overlay-like insertion patterns, and host directives for composing behaviors across components. Trigger this when you need reusable DOM behaviors, to extend element functionality, or to compose behaviors between components. Note: prefer native control-flow directives (@if/@for/@switch) for basic branching and loops rather than creating custom structural directives.

Design QMS features following an MRM-centric architecture. Use when the user asks to "design QMS feature", "build compliance feature", "create MRM section", or needs to ensure features align with ISO 9001 requirements and Management Review Meeting workflows.