Results for “grading”

Evaluates execution transcripts and output files against a list of expectations, assigning pass/fail verdicts with cited evidence and critiquing the assertions themselves.

Grades and classifies evaluation batch results, applying exclusions, diagnosing failure modes, computing pass rates, and generating summary tables for papers.

Grades individual test methods and produces a compact markdown table with a letter grade, score band, and one-line note for each test.

Apply consistent photo adjustments across a set of images so they look like they were edited together, using Adobe tools for color, tone, cropping, and optional selective enhancements.

Provides production-grade techniques for evaluating LLM outputs using LLMs as judges, covering direct scoring, pairwise comparison, bias mitigation, rubric generation, and confidence calibration.

Benchmark-relative quality gate. Scores the draft against the research dossier's beat spec (depth, consensus coverage, evidence) plus AI-tell and voice signals, runs an adversarial read armed with the SERP benchmark, and emits the verdict that gates the pipeline.

Multi-dimensional lead qualification scoring. Evaluates leads against BANT criteria, firmographic fit, behavioral signals, and intent indicators. Outputs qualified/disqualified verdict with detailed reasoning.

Design structured evaluation rubrics for scoring LLM and agent outputs — defining quality dimensions, scoring scales, hard gates, score descriptions, and edge cases. Load when the user asks to create an eval rubric, define evaluation criteria, design scoring dimensions, write an eval spec, or says "what should I evaluate", "design a rubric", "create eval criteria", "define quality dimensions", "evaluation rubric for", "how do I measure quality of". Sub-skill of eval-output orchestrator.

Evaluate LLM outputs systematically — benchmarks, automated metrics, human preference, and regression tracking

Evaluates RAG pipelines using a filesystem-based benchmark with corpus/ and train.json, running evaluate_rag.py to tune retrieval and generation flags and interpret RAGAS metrics.

Defines ideal customer profile filters, scores inbound and outbound leads, and builds a lightweight qualification rubric to sharpen pipeline focus for founder-led sales.

Check post-launch adoption for a published ORD package and recommend iterate, graduate, or archive

Decide whether an article deserves publication through hard checks and skeptical comparison.

Gatekeeping Git/PR by classifying change essence and recommending granularity, naming, and strategy. Use when PR preparation or commit strategy is needed.

Honestly evaluate AI work quality using a two-axis scoring system with mandatory devil's advocate reasoning and cross-session anti-inflation detection.

TRAK: Attributing Model Behavior at Scale

Aesthetic assessment and design scoring across 6 dimensions. Use for UI critique, design review, visual quality assessment, remix suggestions. Activate on "design critique", "aesthetic review", "UI assessment", "visual quality", "design score", "remix this design". NOT for implementation (use frontend-developer), accessibility-only audits (use color-contrast-auditor), or brand identity creation.

Use to staff and schedule QA reviewers for sustainable throughput without grading quality collapsing — fatigue, drift, and false precision from treating reviewers like production agents. Trigger for "how many QA reviewers do we need", reviewer capacity planning, grader fatigue, throughput per reviewer, calibration after long shifts, or scores drifting over the course of a day.

Rank an existing set of context chunks by relevance, diversity, freshness, and utility. Use when retrieval has already produced candidates that must be scored or reranked; use context-retrieval when the source corpus still needs to be searched.

選択肢(A/B/C)の深掘り比較→淘汰→推奨で判断負担を下げ判断の質を上げるスキル。5段階（L1固定3点/L1.5案拡張Diverge・自動/L2評価軸マトリクス/L3複数LLM弁証論/L4過去判断照合）。 「比較して」「深掘りして」「メリデメ教えて」「お勧めは？」「徹底的に」「過去の判断と照合」「前にどう決めたっけ」「/sentaku」等で発火。teian(浅)の深掘り要求を受け取り、brainstorming(深:設計全体)と棲み分け。

Turn a pile of accounts into a stack-ranked priority list with a reason on every row. A layered score (gates first, then an evidence-weighted base rank over the signals you actually have, then bounded boosts for product usage and buyer intent) that stays fair across channels and never scores a blank field as a zero. Built for B2B GTM teams, customizable to your signals and your ICP. Trigger on "score these accounts", "rank by fit", "composite ICP score", "stack-rank my list", "who should I work first", "prioritize these leads", or any multi-signal account qualification.

PASS/FAIL eval rubrics and alignment loops for Sid Orchestra (global). Use when the user says sid evals, @sid-evals, grade this, eval gate, alignment score, or wants to stop AI slop. Works in any workspace; bootstraps EVALS.md from ~/.cursor/skills/sid-orchestra/templates if missing.

Score a UI file's design quality 0-100 against StyleSeed's design language with per-category breakdown, worst offenders, and prioritized fix list.

The single highest-leverage convention in code review is prefixing every comment

Vendor privacy risk tiering methodology for processor management. Covers scoring factors including data volume, sensitivity, transfer locations, certifications, breach history, and control maturity with weighted risk calculation and tier assignment.

Use when designing a course, assessment, or student-feedback workflow and deciding where AI assistance is acceptable, when grading or feedback must stay within academic-integrity and FERPA/KVKK boundaries, or when AI-use rules for students need stating.

Design badge and achievement systems for learning. TRIGGERS - Use when user needs help with badge-system related tasks.

Evaluate and rank agent results by metric or LLM judge for an AgentHub session.

Audits typographic decisions on a screen for scale usage, readability, consistency, and token compliance, providing specific fixes.

Improve existing content and close the learning loop without cannibalizing new-content work.