Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/ryanmaclean/skills.Loading the repository README…
Showing 40 of 112 collected skills.
Run bop card work inside disposable FreeBSD jails on fbryz3070 — golden ZFS dataset cloned per card, nullfs workdir r/w + skills r/o, destroy on exit. Use when the factory needs worker isolation, per-card credential scoping, or sandboxed agent execution on…
Specialized roundtable personas and a phased execution plan for repository issue resolution.
Build, query, and export a context graph of codebase dependencies and relationships.
Unified AI provider quota monitor — Claude Code, Codex CLI, and Gemini CLI OAuth rate-limit utilization bars with reset countdowns, read from local auth files (no API keys). Use when checking quota, rate limits, provider headroom, or why a provider is…
Reference for writing portable /bin/sh scripts targeting ash (Almquist shell) — the default shell on FreeBSD, NetBSD, and TinyBSD fleet nodes. Covers POSIX features, bash incompatibilities, fleet patterns, rc.d service scripts, and UTF-8/Unicode filename…
Apple Silicon optimization for VMs, ML inference, and compression. Covers Metal GPU acceleration, MLX model inference, CoreML conversion, M-series zstd compression, Apple Virtualization Framework, and performance tuning. Use when working on macOS M-series…
Manage Gas Town agent orchestration, bead lifecycle, Kafka messaging, Airflow DAGs, and multi-rig coordination. Use for gt operations, polecat management, stuck beads, Kafka topics, and Tundra Dome workflows.
Manage RigFleet bot configuration — channels, agents, security, and autopilot settings. Use when changing openclaw/RigFleet bot channels, agent assignments, or autopilot behavior.
Surgical code refactoring to improve maintainability without changing behavior. Covers extracting functions, renaming variables, breaking down god functions, improving type safety, eliminating code smells, and applying design patterns. Less drastic than…
Manage webcam streaming from an ATEM Mini connected to Raspberry Pis — USB-C webcam capture (macOS avfoundation, Linux v4l2), HDMI input switching via the ATEM network protocol (UDP 9910), raw-socket Python control, Ansible playbooks, and IRC bot dispatch.…
Autonomous goal-directed iteration applying Karpathy's autoresearch loop — modify, verify, keep/discard, repeat. Use when the user wants a task iterated autonomously, says /autoresearch, /loop, hill-climb a goal, or asks for iterative self-improvement on…
Keep a PR merge-ready by triaging comments, resolving clear conflicts, and fixing CI in a loop. Use when a pull request needs watching, CI keeps failing, review comments are piling up, or the user says babysit this PR.
Datadog unified service tagging best practices — env/service/version core tags, naming rules, and tag standards across APM traces, logs, and metrics. Use when setting up DD_SERVICE/DD_ENV/DD_VERSION, normalizing tags, or auditing tag coverage.
ESP32-C6 RISC-V UART-to-IRC bridge for Tang FPGA console — ESP-IDF v5.3 build, USB flash from studio, WiFi to fleet LAN, plaintext IRC to Ergo, WireGuard support. Use when working on the Tang FPGA serial bridge, ESP32-C6 firmware, or UART-to-IRC plumbing.
Verify a claim against reality — shapes assertions into {subject, expected, probe, evidence, verdict} and runs independent probes. Use when another agent reports X works/ships/is deployed and proof is needed before trusting it, or when the user says verify…
Kubernetes cluster management with k9s, kubectx, and kubectl-ai. Use when switching contexts, inspecting pods, or managing k8s clusters — especially with AI-assisted kubectl.
Architecture review gate for lower-bound systems research. Use before proposing a new queue, scheduler, database, storage layer, FPGA block, RISC-V core, ASIC function, lineage store, runtime, or cross-repo primitive — forces prior-art checks, lowest-layer…
Validate Apple Metal GPU telemetry in Datadog and verify DCGM shim ingestion across workstations. Use when GPU metrics look missing, stale, or inconsistent on Apple Silicon training hosts, or when checking the DCGM-shim pipeline.
Browse and search the MiSTer FPGA forum (misterfpga.org) efficiently via SearXNG cache. Use when looking up MiSTer forum threads, core discussions, or FPGA community answers without hitting the forum directly.
NetBSD package management and Rust build setup on QEMU aarch64 VMs. Use for pkg_add/pkgin on NetBSD ARM64, installing Rust toolchains, or preparing NetBSD build VMs.
Local-first, security-first control center for RigFleet agents — visibility dashboard with readonly defaults, token attribution, collaboration tracing, and safe write operations. Use for RigFleet dashboards, token usage views, or control-center write ops.
Find and test free OpenRouter models — lists available free models, tests availability, and selects the best working model for code generation. Use before spawning agents to pick cost-effective models, or when asked which OpenRouter model to use.
Pre-dispatch quota check showing live utilization across providers (Codex, Auto-Claude, others), recommending the backend with the most headroom — warns above 70%, blocks above 90%. Use before delegating tasks, dispatching agents, or when asked what has…
Automatic verification hook for sub-agent claims — extracts <claims> blocks from agent reports on SubagentStop, verifies each via fleet-eval probes in parallel, and injects verdicts into context. Use when wiring claim verification, debugging the hook, or when…
Manage QEMU virtual machines on the fleet (Apple Silicon + Pi nodes). Use when creating, booting, snapshotting, or debugging QEMU VMs on fleet hosts.
End-to-end QNAP SNMP monitoring triage — Datadog NDM, pup metrics, SSH reachability, UDP/161, and optional snmpwalk checks. Use when QNAP/NAS SNMP metrics are missing, stale, or noisy in Datadog.
Docker operations on qnas (QNAP NAS at 10.0.2.35) — Docker is not in PATH, so use the full container-station binary path for ps, run, exec, logs, pull, and rm over SSH. Use for any container task on the QNAP host.
Patch Rust crates that fail to build on NetBSD aarch64 (ARM64). Use when cargo builds break on NetBSD, when a crate needs a BSD patch, or when cross-compiling Rust for aarch64 NetBSD.
Use the local SearXNG instance (searxng.local) for web searches when WebSearch is rate-limited, geo-blocked, or returns no results — also for fleet-internal search via JSON API.
Run the rest of a /shell request as a literal shell command. Use only when the user explicitly invokes /shell and wants the following text executed directly in the terminal.
Capture and render terminal output as SVG screenshots using termframe — for command output, remote sessions, and OpenCode reporting workflows. Use when asked to screenshot a terminal, render ANSI to SVG, or produce terminal output images.
Hill-climb a skill's prompt/SKILL.md against binary eval assertions using the autoresearch-skills harness. Baseline → generate variants → score each → keep the winner → repeat until plateau or 90%. Use when a skill underperforms, when adding a new skill that…
Manage the Asustor AS6510T NAS (anas, 10.0.3.9) — Docker containers, networking, storage, ADM configuration, and fleet agent deployment. Use for asustor, anas, or AS6510T operations including the /24-vs-/23 route fix.
Qdrant CLI for fleet agents — wraps the Qdrant REST API with built-in embedding via Ollama snowflake-arctic-embed2. Use for vector search, collection browsing, upserts, and purges across fleet Qdrant collections.
Run Deming Plan-Do-Study-Act (PDSA/PDCA) cycles with explicit hypotheses, measurable gates, proof probes, and standard-work updates. Use when driving execution under uncertainty, recovering stalled build lanes, coordinating parallel agents, or reducing repeat…
Enforce safe authentication behavior for GitHub and other CLIs. Use when handling credentials, checking auth status, or debugging agent auth failures so login churn and lockouts never happen.
Prepare and repair Paperclip heartbeat execution by turning vague or stuck issues into executable work packets, decomposing large tasks into assigned subtasks, and posting targeted wake comments that trigger the right agent. Use when runs are failing/idle,…
Operate Paperclip companies and agents safely and deterministically. Use when assigning work, unblocking heartbeat loops, repairing stale in_progress issues, and standardizing API command patterns.
Apply Toyota Way (A3, PDCA, standard work) with TRIZ for MLX model work. Use for planning, decomposition, benchmarking, and continuous improvement across agents.
Apply TRIZ to MLX model-roadmap contradictions where improving one goal harms another (quality vs speed, memory vs context length, novelty vs reliability). Use when deciding which models and quantizations to convert next, what to run in parallel, and what to…
Repository README source: README.md
Read the source README