Agent skills you can check before you install
Public SKILL.md skills from GitHub repositories with 50+ stars and from the vendors themselves, scanned and scored. Install any of them with one command in Claude Code, Cursor, Codex, GitHub Copilot or Gemini CLI.
49,492 skills from 2,391 repositories · updated 16 hours ago
17,846 skills
- meta-evalagentscope-ai/OpenJudgeUse when the user wants to build an evaluation system for an LLM/agent application but doesn't know where to start — they have traces, prompts, RAG pipelines, or nothing at all. Also use when the user mentions evaluation, eval, benchmarking, testing LLM quality, measuring agent performance, assessing RAG accuracy, or wants to compare prompts/models. This skill is the entry router: it asks diagnostic questions then recommends which sub-skill (local workflow) to use next.TestingNo warnings★ 1K↓ —
- 00-academic-routeragentscope-ai/OpenJudgeUse when the user wants help with academic papers or citations but it's unclear which specific workflow fits — reviewing a paper, checking a BibTeX file for fake references, or benchmarking multiple LLMs on reference-recommendation accuracy. Also use when the user mentions paper review, peer review, BibTeX verification, citation checking, reference hallucination, or academic literature accuracy and hasn't specified which of those three tasks they mean. This skill is the entry router for the academic-eval suite: it asks one diagnostic question then routes to the right sub-skill.ScienceNo warnings★ 1K↓ —
- 02-rl-rewardagentscope-ai/OpenJudgeBuild RL reward signals using the OpenJudge framework. Covers choosing between pointwise and pairwise reward strategies based on RL algorithm, task type, and cost; aggregating multi-dimensional pointwise scores into a scalar reward; pairwise tournament reward for GRPO on subjective tasks (net win rate across group rollouts); generating preference pairs for DPO/RLAIF; and normalizing scores for training stability. Use when building reward models, scoring rollouts for GRPO/REINFORCE, generating preference data for DPO, or doing Best-of-N selection.ProductivityNo warnings★ 1K↓ —
- image-to-uihashgraph-online/awesome-codex-pluginsRecreate an image, screenshot, or mockup as a 1:1 UI implementation. Match the reference rather than redesigning it, then render and compare the result to correct visible differences.DesignNo warnings★ 1.5K↓ —Also in 1 other repo
- agm-in-a-boxmohitagw15856/pm-claude-skillsRun a club, PTA, or association AGM that finishes on time and holds up later — the notice and agenda done right, a quorum plan, minutes that capture decisions not conversations, elections without awkwardness, and the follow-up that makes decisions real. Use when a volunteer says 'I have to run the AGM', 'what goes in the agenda', 'nobody comes to our meetings', or 'our elections are a mess'. Produces the notice, agenda, chair's script, minutes template, and quorum rescue plan.PlanningNo warnings★ 1.5K↓ —
- ai-eval-planmohitagw15856/pm-claude-skillsDesign an evaluation plan for an LLM or AI feature before shipping it. Use when asked how to evaluate a prompt/model/agent, set up an eval harness, define quality metrics for an AI feature, or build a regression gate. Produces an eval plan — task definition, datasets, metrics & rubrics, baselines, automated + human evals, a pass bar, and a regression gate.PlanningNo warnings★ 1.5K↓ —
- chart-choicemohitagw15856/pm-claude-skillsPick the chart the data and the point actually need — the question-to-chart mapping (comparison, trend, composition, distribution, relationship), the honesty rules (axes, baselines, dual-axis traps), and the one-chart-one-point discipline. Use when asked what chart should I use, make this data visual, why does this chart feel misleading, or fix this graph for the deck. Produces the chart verdict with its reasoning, the honesty checklist applied, and the labeling that lets the chart travel without its author.DocumentsNo warnings★ 1.5K↓ —
- vuln-triagemohitagw15856/pm-claude-skillsTriage a vulnerability or scanner finding — assess real severity, exploitability, and how urgently to fix. Use when asked to triage a CVE, prioritize scanner/pentest findings, assess a vuln's risk, or decide what to patch first. Produces a triage verdict: CVSS-informed severity adjusted for your context, exploitability, real risk, a fix/mitigation, and an SLA — so you fix what matters, not just what's red.SecurityNo warnings★ 1.5K↓ —
- what-to-askmohitagw15856/pm-claude-skillsGet the five questions that matter before you sign, buy, or agree to anything — the front door to the decoder family, routed by situation. Use when asked what should I ask before signing this, I'm about to buy X what do I check, what questions for the landlord/dealer/contractor/HR, or what am I forgetting. Produces the five highest-leverage questions for the specific situation with why each matters and what a bad answer sounds like, plus the pointer to the full decoder when one exists.PlanningNo warnings★ 1.5K↓ —
- masking-budgetmohitagw15856/pm-claude-skillsTreat neurodivergent masking as a daily energy budget — audit what passing as neurotypical actually costs you, where the spend is worth it, where you can safely drop the mask, and how to plan a heavy-masking day so you don't crash after. Use when someone says 'I'm exhausted from masking', 'work drains me and I don't know why', 'how do I unmask safely', or is autistic/AuDHD/ADHD and burning out socially. Produces a mask-cost audit, a spend/drop map, and a recovery plan. A self-knowledge tool, not a diagnosis or therapy.PlanningNo warnings★ 1.5K↓ —
- evalhashgraph-online/awesome-codex-pluginsQuality and performance evaluation with baseline comparison. Sub-modes: correctness, performance, quality, regression. Outputs PASS/WARN/FAIL per dimension. Use for pre-ship evaluation or regression checks.OtherNo warnings★ 1.5K↓ —
- minutes-x1-send-meetingsilverstein/minutesSend one of the user's own Minutes meetings to their X1 household for review. Use when the user wants a meeting's summary, decisions, action items, and open questions to reach X1 so they can confirm what belongs in their household record. X1 asks the user to approve the send, and nothing reaches the household record until they confirm each item. Never use it for a restricted meeting, a transcript, or someone else's meeting.EngineeringNo warnings★ 1.6K↓ —Also in 1 other repo
- documenthashgraph-online/awesome-codex-pluginsRecord a decision, document existing code, or file a supplied research material. Modes: document decision (ADR, RFC, or rule), document code (spec, doc, guide, or scenario for existing behavior), document research (only when a finished report or one external material is already in hand). Use for 'we decided', 'record this decision', 'document why we chose X', 'make it our standard', 'draft an RFC', 'should we switch to Y' proposals, 'resolve the RFC', 'we accepted the proposal', 'document the auth module', 'capture how the payment system works', reference material, how-to instructions, or a user flow with examples over an existing spec. Name the type inside the subject to skip the type question. Planning a feature or an intended user journey → /archcore:plan. Checking docs against code or docs health → /archcore:review.ScienceNo warnings★ 1.5K↓ —
- store-watchmohitagw15856/pm-claude-skillsHave a proactive agent watch an online shop (Shopify, WooCommerce, Etsy) for what needs the owner: best-sellers about to run out, a spike in refunds or returns on one product, bad reviews, failed orders and checkout problems, with interrupts only for what cannot wait and drafted replies for reviews. Use when asked to monitor my Shopify store, have Duet watch my shop, alert me when stock runs low, watch for bad reviews, or daily store health check. Produces the watch rules, the daily store lines, the interrupt thresholds, and drafted review replies; it never changes prices, refunds orders or publishes anything.FinanceNo warnings★ 1.5K↓ —
- form-explainermohitagw15856/pm-claude-skillsExplain an official form field by field (tax, benefits, immigration, employment, school, health, vehicle or court forms) from a photo, PDF, pasted text or just its name and number: what it is for, what each field means and what to put, the documents needed, the deadline, and where to send it. Use when someone says 'what is this form', 'help me understand form P45 / W-4 / Form 16', 'what do I write in this box', or 'I got this form in the post'. Produces a one-line purpose, a field-by-field guide, a documents checklist, common mistakes, and what happens next.DocumentsNo warnings★ 1.5K↓ —
- update-coghuytieu/COG-second-brainCheck for and apply upstream COG framework updates (skills, docs, scripts) without touching personal contentOtherNo warnings★ 1.3K↓ —Also in 1 other repo
- onboardinghuytieu/COG-second-brainPersonalize COG for your workflow - creates profile, interests, and watchlist files with guided setup (run this first!)ProductivityNo warnings★ 1.3K↓ —Also in 1 other repo
- self-reviewmohitagw15856/pm-claude-skillsWrite a performance self-review that's specific, evidenced, and balanced. Use when asked to write a self-review, self-assessment, or self-evaluation for a performance cycle. Produces a complete self-review — accomplishments mapped to impact and competencies, growth areas owned honestly, and a forward-looking development plan, in the voice of the person being reviewed.EngineeringNo warnings★ 1.5K↓ —
- award-applicationjezweb/claude-skillsWrite compelling award submissions, grant applications, and competition entries. Maps achievements to selection criteria using evidence-based narratives. Use whenever the user is applying for a business award (Telstra, chamber of commerce, industry awards), a grant, a competition, or any submission demonstrating merit against defined criteria.WritingNo warnings★ 1.1K↓ —
- tos-decodermohitagw15856/pm-claude-skillsDecode a terms of service or privacy policy into what you're actually agreeing to, ranked by real-world impact. Use when someone asks 'what am I agreeing to', 'decode this privacy policy', 'is this ToS bad', or 'should I click accept'. Produces a ranked findings table with a 'should I care?' verdict per finding, covering data resale, arbitration and class-action waivers, unilateral changes, content licenses, and what deletion really means.OtherNo warnings★ 1.5K↓ —
- clip-factorymohitagw15856/pm-claude-skillsTurn one long video, podcast, or stream transcript into 8-12 short-form clips — each with a hook line, cut timestamps, captions, and a platform note for TikTok/Reels/Shorts — plus an honesty gate that kills clips that misrepresent the source. Use when someone says 'clip this podcast', 'make shorts from my video', 'what's clippable here', or runs a clipping side hustle. Produces a ranked clip sheet ready for an editor or a clipping app.MediaNo warnings★ 1.5K↓ —
- analyze-performance-revenue-joingooseworks-ai/goose-skillsStep of the analyze-performance playbook: joins cleaned store or CRM numbers with ad spend to build the business-truth table, MER, new-customer cost and the overclaim ratio. Run analyze-performance instead; not for direct use.MarketingNo warnings★ 1.3K↓ —
- validatehashgraph-online/awesome-codex-pluginsRun the profile-aware DocFlow documentation readiness gate. Always blocks broken links, stale maps, and missing headings; applies native naming and section rules strictly to new scaffolds while treating established adopted structures as cleanup guidance.OtherNo warnings★ 1.5K↓ —
- oma-qafirst-fluke/oh-my-agentReview changes for correctness, security, accessibility, and performance. Use for scoped quality reviews or verification plans; bug fixes use oma-debug.TestingNo warnings★ 1.4K↓ —Also in 2 other repos
02-rl-reward
agentscope-ai/OpenJudge
ProductivityUpdated 2 days ago
From the author
Build RL reward signals using the OpenJudge framework. Covers choosing between pointwise and pairwise reward strategies based on RL algorithm, task type, and cost; aggregating multi-dimensional pointwise scores into a scalar reward; pairwise tournament reward for GRPO on subjective tasks (net win rate across group rollouts); generating preference pairs for DPO/RLAIF; and normalizing scores for training stability. Use when building reward models, scoring rollouts for GRPO/REINFORCE, generating preference data for DPO, or doing Best-of-N selection.
Install
npx skills add agentscope-ai/OpenJudge --skill 02-rl-reward -a claude-code
Installs to .claude/skills/02-rl-reward/ · Each command fetches the skill straight from GitHub with the open skills CLI.
Download .zip
Byte-identical to the scanned commit, with the repository’s licence and a SOURCE.txt.
What it can touch
Instructions onlyUses the networkLicence: APACHE-2.0
Safety checks on SKILL.md
No warnings.
Why 88
Security100
Provenance76
Maintenance100
Licence100
Adoption49
Files (3)
- SKILL.md
- pairwise.md
- pointwise.md