DevOps & Infra

ryan-brosas Bundle

Use when porting ell (language model programming) — LMP decorators, _lstr origin tracing, lexical-closure versioning, provider SPI, SQL/alembic store, evaluations, studio; prompt-to-program frameworks that version prompts as code (lexical closures + content hashes), trace which model call produced which text fragment through arbitrary string mutations, adapt one call pipeline to many vendor APIs behind a validation-enforced provider interface, or persist/version LLM programs and invocations in SQLite/Postgres with alembic. Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.

ryan-brosas Bundle

Use when building an agent harness's safety layer AND context plane: pure approval classification with timeout auto-answer, budget-window spend caps, tool repetition breakers, model-driven persona routing, promise-gated late config, stack-driven request loops, shadow-git checkpoints, non-destructive condense/truncate history management, dual-log persistence with rewind reconciliation, execution-time tool gating, native tool-call stream parsing with fail-fast validation, fuzzy SEARCH/REPLACE diff application, .rooignore access control, mention-to-context pipelines, multi-instance task history stores, budgeted prompt-cache breakpoint placement, bounded ripgrep search transport, tree-sitter definition-outline extraction over a lazy WASM grammar loader, AND a whole-file-mined workspace→Qdrant code-index RAG pipeline (scanner/parser/watcher/orchestrator, 8-embedder batching kernel, pathSegments filter grammar, in-collection indexing markers, dimension-mismatch recreation).

ryan-brosas Bundle

Use when porting Isso's self-hosted comment-server internals: SQLite comment/thread storage, moderation modes, guard rate limits, signed edit/moderation tokens, notification fanout, sanitizer pipeline, or the multi-mixin WSGI deployment core.

ryan-brosas Bundle

Use when building or porting a DI container, module system, provider lifecycle, bootstrap pipeline, request-scoped injection, route registration/URL composition, validation pipes, or a boot log buffer — NestJS's core injector + router + pipe kernel as source-confirmed contracts.

ryan-brosas Bundle

Use when porting or reimplementing visualization primitives from visx (@visx/*) — scale operator pipeline, chart keyboard-nav state machine + ARIA tree, brush extent FSM, zoom matrices, responsive/text/spring primitives: d3-scale config application with correct operator ordering, keyboard-navigable + screen-reader-accessible charts, pixel-to-domain brush selection with ordinal support, cursor-anchored zoom transforms, parent-size responsiveness without layout loops, SVG text measurement/wrapping, animated line enter/exit trajectories, and referentially-stable data/accessor plumbing.

synthetic-sciences Bundle

This skill should be used when users want to run any workload on Hugging Face Jobs infrastructure. Covers UV scripts, Docker-based jobs, hardware selection, cost estimation, authentication with tokens, secrets management, timeout configuration, and result persistence. Designed for general-purpose compute workloads including data processing, inference, experiments, batch jobs, and any Python-based tasks. Should be invoked for tasks involving cloud compute, GPU workloads, or when users mention running jobs on Hugging Face infrastructure without local setup.

synthetic-sciences Bundle

Write comprehensive clinical reports including case reports (CARE guidelines), diagnostic reports (radiology/pathology/lab), clinical trial reports (ICH-E3, SAE, CSR), and patient documentation (SOAP, H&P, discharge summaries). Full support with templates, regulatory compliance (HIPAA, FDA, ICH-GCP), and validation tools.

synthetic-sciences Bundle

Submit and manage protocols on Ginkgo Bioworks Cloud Lab (cloud.ginkgo.bio), a web-based interface for autonomous lab execution on Reconfigurable Automation Carts (RACs). Use when the user wants to run protein expression and purification (cell-free, E. coli, or Pichia), HiBiT or A280 or LabChip quantification, IVT mRNA/circRNA synthesis, thermal shift / developability assays, Echo-MS enzyme or analyte methods, SPR target onboarding, fluorescent pixel art, or otherwise interact with Ginkgo Cloud Lab services. Covers protocol selection, input preparation, pricing, and ordering workflows.

synthetic-sciences Skill

Safely inspect and operate TensorPool GPU clusters and jobs using the current tp CLI, with explicit approval before any billable or destructive action.

synthetic-sciences Skill

Operate GiveMeANode GPU nodes, batch jobs, storage, and rollout sandboxes through its connected MCP server with bounded spend, durable recovery, and explicit approval for paid or destructive actions.

synthetic-sciences Skill

Safely inspect and operate Lambda Cloud GPU instances through the documented Cloud API and SSH, with explicit approval before billable or destructive actions.

synthetic-sciences Skill

Serverless inference, fine-tuning, embeddings, image generation, and batch processing on 200+ open-source models via an OpenAI-compatible API. Use when you need fast, cost-effective access to open-source LLMs without managing infrastructure.

synthetic-sciences Bundle

Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Use when training models >1B parameters, need maximum GPU efficiency (47% MFU on H100), or require tensor/pipeline/sequence/context/expert parallelism. Production-ready framework used for Nemotron, LLaMA, DeepSeek.

synthetic-sciences Skill

Fast inference and fine-tuning platform with serverless and on-demand GPU deployments. OpenAI-compatible API for chat completions, embeddings, function calling, vision, and structured output. Supports SFT, DPO, and RL fine-tuning. SOC2 + HIPAA compliant.

synthetic-sciences Bundle

Comprehensive geospatial science skill covering remote sensing, GIS, spatial analysis, machine learning for earth observation, and 30+ scientific domains. Supports satellite imagery processing (Sentinel, Landsat, MODIS, SAR, hyperspectral), vector and raster data operations, spatial statistics, point cloud processing, network analysis, cloud-native workflows (STAC, COG, Planetary Computer), and 8 programming languages (Python, R, Julia, JavaScript, C++, Java, Go, Rust) with 500+ code examples. Use for remote sensing workflows, GIS analysis, spatial ML, Earth observation data processing, terrain analysis, hydrological modeling, marine spatial analysis, atmospheric science, and any geospatial computation task.

synthetic-sciences Bundle

Execute Hugging Face Hub operations using the `hf` CLI. Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs on HF infrastructure. Covers authentication, file transfers, repository creation, cache operations, and cloud compute.

synthetic-sciences Bundle

End-to-end molecular docking pipeline. Target preparation, pocket detection, protein-ligand docking (DiffDock/Vina), scoring, interaction analysis, and pose ranking.

synthetic-sciences Bundle

Complete flow cytometry analysis pipeline. FCS file handling, compensation, manual/automated gating, immunophenotyping, CFSE proliferation analysis, cell cycle analysis (Dean-Jett-Fox), and apoptosis assays. Extends flowio with analytical workflows. For raw FCS parsing only use flowio.

synthetic-sciences Bundle

Calculates training costs for Tinker fine-tuning jobs. Use when estimating costs for Tinker LLM training, counting tokens in datasets, or comparing Tinker model training prices. Tokenizes datasets using the correct model tokenizer and provides accurate cost estimates.

synthetic-sciences Bundle

Build training datasets for LLM specialization from production data, frontier model distillation, and synthetic bootstrapping. Use when formatting production logs into SFT data, distilling from frontier APIs, or preparing data for fine-tuning. Covers JSONL formatting, data quality validation, deduplication, and train/eval splitting.

synthetic-sciences Bundle

This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.

dojogenesis Skill

Produces a scout document: identifies a tension, generates 3-5 distinct routes with tradeoffs, synthesizes a recommendation, and proposes a first action. Lean single-file variant with no scout-to-spec pipeline scaffolding — full version: `strategic-thinking:strategic-scout`. Use when: 'scout this', 'strategic scout', 'explore options for', 'what are the routes'.

wyre-ai Skill

Axcient x360Recover vaults: Private vs Cloud vault types, storage capacity, and the connectivity-loss alert threshold — including axcient_set_vault_threshold, the one tool in this plugin that changes alerting configuration rather than just reading it.

wyre-ai Skill

Axcient x360Recover clients: the health_status model, client_code, the three-way protected-system counter breakdown (appliance_based, d2c, cloud_archive), and minting direct-to-cloud agent enrollment tokens.

wyre-ai Skill

Axcient x360Recover protected devices (servers and workstations): SERVER/WORKSTATION typing, the current/previous health-status model, local vs vault vs cloud usage and recovery-point timestamps, AutoVerify screenshot verification, and restore points.

wyre-ai Skill

HubSpot support tickets: core fields, default pipeline stages, priority levels, MSP ticket categories, SLA timestamp properties, and associating tickets with contacts, companies, and deals.

wyre-ai Skill

The Nutanix networking read surface: `networking_execute` for AHV network configuration (subnets, VPCs, vSwitch, BGP — coverage varies by artifact set, down to a minimal capabilities/AWS-VPC slice) and `microseg_execute` for Flow Network Security policies, service groups, and address groups.

wyre-ai Skill

CRM pipeline health assessment against whatever CRM is discovered through the gateway: stage-velocity norms derived from closed-won deals, activity-based stalled-deal detection, raw and quality-adjusted pipeline coverage against a revenue target, and the CRM-less degradation rule (report nothing rather than fabricate figures).

wyre-ai Skill

MSP proposal workflows in PandaDoc: proposal types and typical values (MSA, SOW, hardware quote, project proposal, security assessment, cloud migration), the standard MSP content-token set, managed-services and hardware pricing-table structures, mapping document statuses to sales-pipeline stages, and pipeline and stale-proposal tracking.

wyre-ai Skill

Three-way seat reconciliation per client per SKU: seats provisioned in a cloud marketplace (Pax8, Sherweb) vs. seats billed in accounting or PSA billing vs. seats actually deployed in the tenant (microsoft-graph or CIPP). Covers the finding class for each pairwise mismatch, disabled users holding paid licenses, intentional-buffer caveats, and how to degrade to a two-way check when an axis isn't connected.

wyre-ai Skill

Spanning Cloud Backup REST API fundamentals: admin-email + API-token auth, the per-platform endpoint surface (M365, Google Workspace, Salesforce), the user/license model, backup status queries, and restore operations.

wyre-ai Bundle

Liongard's inspection pipeline: inspector templates and their credential and agent requirements, launchpoint configuration that binds inspector + environment + agent + credentials + cron schedule, on-demand inspection runs and their status lifecycle, and the failure modes behind failed runs.

wyre-ai Bundle

Reconciling cloud marketplace subscriptions (Pax8) against accounting invoices (Xero, QuickBooks Online): the matching strategy, billing gaps, unbilled subscriptions, and margin discrepancy analysis.

wyre-ai Skill

Onboarding the azure-mcp connector in the WYRE MCP Gateway: Azure service-principal registration, the tenantId/clientId/clientSecret triple, least-privilege Reader-tier RBAC assignments, the gateway's read-only namespace allowlist, and connection verification and failure modes (expired secret, missing role assignment).

wyre-ai Skill

Inforcer MCP fundamentals: the gateway X-Inforcer-Region / X-Inforcer-Api-Key headers, the region-based base URL and upstream Inf-Api-Key header, the /beta/ route prefix, the {success,message,errors,data} response envelope, continuationToken pagination, and the integer Client Tenant ID vs Azure AD GUID gotcha. Every other Inforcer skill assumes these patterns.

wyre-ai Skill

Assembling a blameless postmortem grounded in systems of record: timeline reconstruction merging the incident tool's event log with correlated observability anomalies (Sentry error spikes, Datadog/Grafana metric anomalies, including precursor signal that predates formal detection) and deploy history from connected platform connectors, all normalized to one stated timezone; the root-cause versus contributing-factor distinction and the test for telling them apart; and how to label a root cause that is still only a hypothesis.