Skill UI

Apply Michael T. Nygard-inspired production reliability rules when designing failure handling, timeouts, retries, circuit breakers, observability, or deployment behavior.

Combines welfare readouts such as weight loss, temperature, clinical scores and biomarkers into a single RELSA severity score per animal per day, with kernel-density attention/danger zones and ARIMA-based forecasting of upcoming humane endpoints for 3Rs refinement reporting.

This skill helps maintainers verify reported bugs in the mcp-toolbox project. It involves analyzing issue reports, checking versions and configurations, and attempting local reproduction to determine if the bug is real, fixed, or misconfigured. The process requires citing specific evidence like file paths and commit SHAs to support verdicts such as confirmed, already-fixed, or client-side issues. It ensures efficient bug triage without unnecessary infrastructure setup.

Sweeps googleapis/mcp-toolbox for issues and PRs with no real activity in N days, classifies each silence as the author's, maintainer's, or nobody's, and drafts nudge or close comments for maintainer review only.

Generate and cut videos against a distilled style pack: plan takes, generate with fal, apply measured LUT grading, edit at cadence, weave footage, composite overlays, mint 3D props, and verify results numerically.

Apply pragmatic programming rules inspired by Hunt and Thomas to improve engineering judgment, feedback loops, DRY knowledge, orthogonality, automation, and adaptability. Use this skill for design and code decisions requiring pragmatic engineering tradeoffs.

TLC Discover transforms unshaped ideas into verdicts and design documents covering decisions, flows, and contracts. It helps teams research problems, understand feasibility, and decide whether to build before planning how. This tool prevents premature solution proposals and ensures concrete values for decisions, creating a plan anyone can follow without attending the original discussion.

This skill extracts a checklist from planned work, builds the implementation, and verifies every check with an independent verifier. It is used when the user requests to extract a checklist, build a ticket, or implement a spec. It ensures rigorous testing and proof-based assertions, preventing self-reporting of completion. It involves extracting checks, building the code, and proving each one independently to guarantee quality.

Turn a PRD, design doc, RFC, or thread into actionable tasks. Finds observable slices, grounds them in code, writes intent and criteria, and flags unresolved gaps. Use for task planning, not discovery or implementation.

A four-step workflow (plan, checks, build, verify) that freezes observable obligations instead of choreographing tasks. One human-reviewed plan with EARS criteria and one-way doors, proof-backed checks, then an independent fresh Verifier over the feature commits.

Work with Outpost Bio's open microbiome foundation models: embed taxonomic profiles, fine-tune Waypoint checkpoints, benchmark on Compass, pretrain on Atlas, and convert MetaPhlAn, Kraken2, QIIME 2, or MGnify tables via the waypoint CLI.

Apply Michael Feathers-inspired rules to safely modify hard-to-test legacy code using characterization tests, seams, sprout methods, or wrap methods. Ideal for risky changes, missing tests, dependency breaking, and incremental risk reduction in software maintenance.