769 skills · sorted by relevance
Build deterministic forecast scripts with Earth2Studio (model, data source, IO, inference). Do NOT use for ensemble, diagnostics, data-only fetch, or install.
The English summary is being prepared. Showing the original description.
Build text tokenizers that split multilingual text into small units for AI models. Learn how to train SentencePiece on raw text, configure BPE or Unigram splitting methods, and convert text to token IDs and back.
Classify outstanding pull request CI failures, automated checks run on code changes, using logs and changed files. Produce JSON decisions recommending a rerun, a comment, or no action, and identify specific nondeterministic tests when evidence supports it.
Check a paper's numerical consistency using fixed rules, then record blocking or warning decisions and outstanding findings. The workflow preserves a traceable correction history, while explicitly leaving broader semantic review unavailable in its Codex mode.
Adjust how autopilot vehicles behave in CARLA, including speed, following distance, traffic light responses, and lane changes. Apply settings across traffic or to individual vehicles, and configure seeds and synchronization for repeatable traffic runs.
Build text tokenizers, which split text into units for AI models, directly from raw multilingual text. Learn custom vocabularies with BPE or Unigram, convert text to pieces or numeric IDs, and decode the results.
Upload local photos, images, or screenshots to Vercel Blob and receive public URLs for sharing or embedding in pull request comments and documents. Supports multiple files, custom path prefixes, and deterministic URLs.
Review a pull request using repository checks, verified code references, and current official documentation. Assess correctness, security, structure, and unnecessary code, then post one consolidated GitHub review with inline findings and an evidence-based verdict.
Assemble a creator picture-in-picture product-listicle ad from a config — the creator stays FULL-FRAME the whole beat (voice plus lips generated together per beat, no separate VO, no cut to a full-frame product shot), and on each product beat three persistent overlays ride on top for the WHOLE beat — a title pill top-center, the DEMO in a rounded PiP window top-right (the brand's real UGC clip MUTED, or for a no-UGC brand the product's own autocropped UI still / screen-recording sized to fill the window), and a bottom product card (rounded thumbnail plus 'N · CATEGORY' small-caps plus product NAME in a serif face). Hook plus CTA beats are the creator full-frame with the title pill only. Assembly builds ONE full-1080x1920 transparent overlay PNG per beat, overlays it on the creator clip (cover-scaled to 1080x1920) for the whole beat keeping the native audio, concats all beats, then burns captions LAST as timed PIL PNG overlays (this ffmpeg has no libass) timed deterministically from the known per-beat script. This is the FREE deterministic assembly stage (overlay-PNG build plus cover-scale composite plus concat plus PIL-PNG caption burn); the creator anchor and the N native talking clips come from create-image-fal (Seedream v5 Pro) and create-video-fal (Seedance 2.0). Use for the creator-pip-listicle format.
The English summary is being prepared. Showing the original description.
Render a 'model comparison grid' video from a config — a fal-style "same prompt, N contenders" showcase — a dark real-DOM stage where per beat a monospace prompt fades in centered, docks to a small top strip, then a labeled 2-4 panel grid (static images OR muted video clips, mixable per cell) staggers in and holds for comparison, plus a minimal end card — frame-stepped via Playwright (video cells are frame-seeked deterministically) and encoded with FFmpeg. Deterministic assembly, FREE (cell media comes from create-image-fal / create-video-fal, music from create-music-elevenlabs), text stays pixel-crisp. Use for the model-comparison-grid format.
The English summary is being prepared. Showing the original description.
Build the deterministic PIL pill overlays for a 'perfect-score + proof-points' UGC video ad (white 10/10 score header with a medal + orange sub with a finger-down + 3-4 green-check proof pills) and composite them onto a base clip in a diagonal L->R->L->R cascade, then mux the music bed into master-final.mp4. Config-driven (config.json), 1080x1920 9:16, FREE and deterministic (no paid calls, text stays pixel-crisp). Use for the overlay-proof-points format; the base clip + music come from separate paid capabilities.
The English summary is being prepared. Showing the original description.
Assemble an editorial-motion podcast-clip ad from a config — a real clipped podcast MP3 carries the narrative while N flat limited-palette editorial-illustration keyframes (one look pack) are animated NOT by generative i2v but by DETERMINISTIC ffmpeg ken-burns (zoompan) + hard cuts (no crossfades, which expose geometric drift), each beat snapped to its spoken line, the real audio muxed, Whisper-driven captions burned only mid-sentence, and closed on a PIL brand end card — never AI-rendered text. This is the FREE deterministic assembly stage (ffmpeg ken-burns + hard concat + audio mux + captions + end card); the real audio is clipped from source and the keyframes come from create-image-fal. Use for the editorial-motion-podcast format.
The English summary is being prepared. Showing the original description.
Create original, competition-ready commercial mediation problems and confidential packets with iterative validation. Use when a user provides a topic, industry, dispute sketch, or existing general-information packet and wants a public packet plus confidential packets for both sides—or one named side. Build a narrow but workable settlement corridor; verify factual and numerical consistency, information asymmetry, balance, originality, and commercial feasibility; and revise until deterministic and semantic exit criteria pass. Do not use merely to summarize an existing packet or to advise parties in a real dispute. Requires local file access; Python 3 is recommended for bundled deterministic checks, with disclosed manual fallbacks when unavailable.
The English summary is being prepared. Showing the original description.
Instruments programs to record execution information for deterministic replay debugging. Use when debugging hard-to-reproduce bugs (race conditions, timing issues, intermittent failures, heisenbugs), reproducing production failures, or analyzing complex execution sequences. Records non-deterministic events (I/O, threading, randomness, time) to enable exact replay of program executions. Supports Python, JavaScript, Java, and C/C++ with both custom instrumentation and existing replay tools.
The English summary is being prepared. Showing the original description.
Instruments programs to capture execution traces specifically for reproducing reported bugs, enabling consistent replay and diagnosis of failures. Use this skill when you need to reproduce a bug, capture execution traces for debugging, instrument code to record program behavior, generate replay scripts for bug reproduction, diagnose hard-to-reproduce failures, or perform deterministic replay of program execution. Triggers when users ask to instrument code for tracing, capture execution traces, reproduce bugs, generate replay scripts, or enable deterministic debugging.
The English summary is being prepared. Showing the original description.
Identifies non-deterministic or unreliable tests through static code analysis and test result analysis. Use when Claude needs to find flaky tests, analyze test reliability, or investigate intermittent test failures. Supports Python (pytest, unittest) and Java (JUnit, TestNG) test frameworks. Trigger when users mention "flaky tests", "intermittent failures", "non-deterministic tests", "unreliable tests", or ask to "find flaky tests", "analyze test stability", or "why tests fail randomly".
The English summary is being prepared. Showing the original description.
Inspect an unfamiliar repository, interpret a broad Markdown user journey at runtime, operate the real product through its supported web, API, CLI, desktop, or mobile surface, and produce an auditable pass, fail, or blocked judgment with screenshots, logs, recordings, and a step timeline when available. Use when asked to run smoke tests, validate an end-to-end user journey, dogfood a product, test a release candidate, or execute a non-deterministic workflow that may include authentication or human checkpoints. Do not use as a substitute for deterministic unit, integration, or scenario tests.
The English summary is being prepared. Showing the original description.
Thesis report synthesis — turn a thesis's markdown artifacts into its letter-size thesis-report.html, then OPTIMIZE it against real page renders. Four-step loop — pack (deterministic Python bundle of all sources + report/metrics.json), author (the LLM writes report/content.html — narrative, KPI tiles, badges, timeline, tables and citations are LLM judgment), assemble (deterministic validation, template injection, Q50 pins, Q47 overflow gate), render + optimize (Chrome headless per-page PNGs — the LLM reads them and iterates until every page is visually clean). Spec 046 Q46–Q50.
The English summary is being prepared. Showing the original description.
Goal-driven E2E testing where a browser agent (Playwright MCP / computer-use) reads a natural-language goal and explores the app via the accessibility tree to assert outcomes — no pre-written script. Covers when intent-driven beats scripted, making agent runs deterministic (pinned model, temperature 0, seeded data, bounded steps, explicit success assertion, snapshot-not-pixel), cost/latency control, the accessibility-tree-first interaction model, CI gating, and graduating a stable run into a scripted Playwright test. Use when: "agentic browser test," "goal-driven browser test," "let an agent explore the app," "natural-language E2E," "browser agent smoke test," "Playwright MCP test." Not for: Writing/maintaining deterministic scripted Playwright tests — that is playwright-automation. Testing your product's OWN LLM features — that is ai-system-testing. Related: playwright-automation, ai-system-testing, exploratory-testing, test-reliability, qa-project-context.
The English summary is being prepared. Showing the original description.
Test AI/LLM features that ship in your product. Covers prompt regression testing, response quality evaluation, tool-call validation, hallucination and RAG grounding checks, nondeterministic-output strategies, red-team/safety scans, eval frameworks, and agent-as-target injection (indirect injection via tool output / RAG / scan reports, self-propagating payloads, data exfiltration via an agent) plus a bundled detector for untrusted content. Use when: "test our LLM feature," "prompt regression test," "eval framework," "hallucination test," "RAG grounding," "nondeterministic output," "AI feature testing," "red-team our chatbot," "indirect prompt injection," "agent reading untrusted tool output," "production AI quality." Not for: using AI to generate your own test code — use ai-test-generation. Not for: classifying CI failures with AI — use ai-bug-triage. Not for: EU AI Act / GDPR conformity of an AI feature — use compliance-testing. Not for: canary/flag rollout of an AI feature — use testing-in-production. Related: ai-test-generation, ai-qa-review, api-testing, compliance-testing, security-testing, risk-based-testing, test-data-management.
The English summary is being prepared. Showing the original description.
Turn a vague bug report into a VERIFIED minimal reproduction and then a failing regression test, agent-driven end to end. Covers extracting the implicit repro from a thin report (env, build, steps, data), the reproduce-minimize-isolate-capture loop, git bisect to find the introducing commit, building a deterministic minimal repro (fixed seeds, frozen time, stubbed network), writing the failing regression test BEFORE the fix (red) and confirming the fix flips it green, and writing repro evidence back into the ticket. Distinguishes flaky-not-reproducible from environment-specific. Use when: "reproduce this bug," "minimal reproduction," "repro steps," "find the commit that broke it," "git bisect," "make the repro deterministic," "write a failing test for this bug," "regression test for a defect," "can't reproduce this bug." Not for: Classifying/deduplicating/severity-routing existing failures without reproducing them — that is ai-bug-triage. Generating tests from specs rather than from a defect — that is ai-test-generation. Related: ai-bug-triage, ai-test-generation, test-reliability, systematic-debugging, qa-project-context.
The English summary is being prepared. Showing the original description.
End-to-end substantive-integrity forensic sweep of a research paper (especially autoresearch / AI-Scientist-style output). Orchestrates the whole pipeline: ingest (arxiv-id | pdf | dir → working dir + pdftotext for L0) → /evidence-ledger (artifact manifest + observability level L0/L1/L2 + span-anchored claims.json) → fan out the integrity auditor skills (consistency, citation, baseline, experiment, presentation, proof-derivation, eval-design — each reads the ledger, emits span-anchored findings) + the zero-verdict-weight AIS writing-style track → advisory memos (/adversarial-case-builder + /novelty-duplication-advisory, no verdict weight) → deterministic tools/adjudicate_findings.py (--ledger REQUIRED) → reviewer-ready Integrity Forensics Report. Cross-model (fresh codex per dimension) and reviewer≠adjudicator: the model proposes findings, the deterministic adjudicator decides the verdict. Observability-aware, detect-only, never an opaque AI-text classifier (a separate zero-weight AIS section lists AI writing-style impressions, never moving the verdict). Triggers: "anti-autoresearch", "integrity audit this paper", "forensic review", "audit a submission", "审一篇投稿的诚信".
The English summary is being prepared. Showing the original description.
Maximum control over AI image generation — write structured VGL (Visual Generation Language) JSON that explicitly controls every visual attribute. Define exact object placement, lighting direction, camera angle, lens focal length, composition, color scheme, and artistic style as deterministic JSON instead of ambiguous natural language. Use this skill when you need reproducible image generation, precise control over scene composition, or want to convert a natural language image request into a structured JSON schema for Bria FIBO models. Triggers on requests for structured prompts, controllable generation, VGL JSON, deterministic image descriptions, or Bria/FIBO structured_prompt format.
The English summary is being prepared. Showing the original description.
This skill should be used when the user asks to "recalculate debt", "利息制限法で引き直し", "引き直し計算", "過払金", "過払い金", "サラ金", "貸金業者との取引履歴から引き直し", or wants deterministic recalculation of loan transactions under the Interest Rate Restriction Act.
The English summary is being prepared. Showing the original description.