Loading...
Loading...
Found 2,404 Skills
Performs ARA Seal Level 2 semantic epistemic review on Agent-Native Research Artifacts, scoring six dimensions (evidence relevance, falsifiability, scope calibration, argument coherence, exploration integrity, methodological rigor) and producing a constructive, severity-ranked report with a Strong Accept-to-Reject recommendation. Use after Level 1 structural validation passes, when an ARA needs an objective epistemic critique before publication or release.
Judge whether tests catch realistic regressions and recommend what to keep, change, or remove.
Build or debug Roblox NPC navigation and behavior. Use for PathfindingService, modifiers and links, blocked paths, patrol or chase logic, streaming, and scaling NPC counts.
Motto: Style counts. Context matters. Story trumps numbers.
Use when you want rubric based LLM quality scoring on generated outputs; pair with addon-deterministic-eval-suite.
Ensure WCAG AA accessibility and readability
Produce a scored heuristic critique of the UI using Nielsen's 10 + 6 design laws + optional persona walkthroughs. Outputs a machine-parseable scorecard plus a 0-100 UsabilityScore (the judged companion to the deterministic UICraftScore). Invoke when the user asks for heuristic on their UI, or mentions 'heuristic' alongside design / UI / frontend work.
Run an iterative code-improvement hill-climb against real Datadog LLM-Obs data, locally, with Claude Code as the agent. Establishes a baseline eval, makes one focused change, re-scores with the same harness, keeps the change if it improves the score in the goal's direction (labeling within-noise gains tentative), and repeats. Use when the user says "run an auto experiment", "hill-climb this code", "iteratively improve X and measure the delta", "optimize this prompt/file against my traces", "auto-optimize against LLM-Obs", or wants the local equivalent of the auto_experiments worker. Works from a local dataset file, an ml_app, a dataset_id, or a list of trace_ids.
verify when running best-of-N, choosing among multiple attempts, a 1-10 or yes/no judge would collapse close traces, comments claim success without a run, or a live attempt may have stalled.
Run a structured usability audit of any UI against Nielsen's 10 heuristics. Produces a severity-rated, prioritized issue list — each finding tagged with the violated heuristic and a concrete fix. Use when auditing a screen, flow, or component for usability problems, or as a validation step after generating UI.
Score a game's combat skills and spells out of 10 with blind judges, then raise the weakest to a bar, for timing and readability (does a skill read as itself, does the blow land when it lands) and for art (detail, light, trails, impact, art direction, judged against a reference such as Diablo IV). Covers filming casts on a stopped, stepped clock, contact strips and close-ups, anonymised two-judge rounds, the levers that moved scores (hero props made from a painted concept with image-to-3D, own schedules and keyed moves, painted additive textures, camera-facing ribbons, light outside the body), the traps that cost points, asset and load budgets, and shipping each skill as its own version. Use when the user asks to score, review, improve, polish or rebuild combat skills, spells or attacks; says a skill "doesn't make sense", "looks bad" or should look like another game; names a target score; or starts a "next skill" thread. Project specifics (review hooks, harness settings, where each skill's code lives, scores so far) go in an optional local references/<project>.md next to this skill.
Turn wiki golden pages into a CI-blocking eval suite with a cheap→expensive assertion cascade plus a committed regression baseline. Goldens are md files under llmwiki/wiki/sources/evals/ with YAML frontmatter (input, expected, optional rubric, asserts). The engine runs deterministic tier-1 asserts (equals/contains/regex/is-json/is-sql-ish) in pure code, an optional tier-2 similarity, and reports tier-3 (LLM-rubric judge) as the quarantined adapter — never calling a model. Candidate outputs come from --outputs outputs.json so the suite is fully runnable without any model. Trigger when the user says "wikieval", "eval suite for the wiki", "regression gate for goldens", "assertion cascade", "block CI on eval drop", or invokes /wikieval.