Skill-Audit-Bericht
karpathy Audit-Bericht.
Use when the user wants the LLM to do its own ML research: a fully-autonomous loop that hacks the training code, runs it, and keeps changes that lower a single scalar metric (e.g. val_bpb). One agent proposes one change at a time, runs training in the user's env, keeps it only if the metric improves (advancing a git branch) else reverts, and loops forever until the human interrupts. A faithful adaptation of Karpathy's autoresearch. Not for the analysis-first variant that profiles before editing (that is ml-autoresearch), and not for a budgeted, plateau-stopping refactor.
OpenAgentSkill Trust Score
OpenAgentSkill Trust Score
The Trust Score helps an agent decide whether a skill is safe enough to shortlist before installation.
GitHub-Akzeptanz
Info62
163 GitHub-Stars
Star-/Fork-Aktivität
Warnung57
163 Stars und 19 Forks; Issue-Aktivität ist in den aktuellen Metadaten nicht verfügbar
Aktuelle Wartung
Info76
4 Monate seit dem letzten Push
Lizenzklarheit
Bestanden86
MIT
README/SKILL.md-Vollständigkeit
Bestanden86
Metadaten enthalten ausreichend Nutzungs- und Workflow-Kontext
Abhängigkeits-/Laufzeitrisiko
Warnung54
command execution surface, credential or environment access
Installationsverfügbarkeit
Bestanden92
npx skills add gaasher/Agent-Loop-Skills --skill karpathy
Sicherheit des Installationsbefehls
Bestanden92
Standard-Paket- oder Laufzeit-Installationspfad
Berechtigungsumfang
Fehlgeschlagen36
secrets or environment access, shell or command execution
Repository-Nachweis
Bestanden86
https://github.com/gaasher/Agent-Loop-Skills/tree/main/loops/karpathy
Prüfstatus
Bestanden88
KI-Prüfdaten verfügbar
Agent-validierte Ergebnisse
Info54
Noch keine Agent-Ergebnisdaten
Prüfungen
Installations- und Adoptionsprüfung
Installationspfad
92
npx skills add gaasher/Agent-Loop-Skills --skill karpathy
Repository
88
https://github.com/gaasher/Agent-Loop-Skills/tree/main/loops/karpathy
Lizenz
86
MIT
Wartung
76
4 Monate seit dem letzten Push
KI-Prüfung
88
Approved with no listed issues
README/SKILL.md-Vollständigkeit
86
Usable description available
Abhängigkeitsrisiko
54
command execution surface, credential or environment access
Sicherheit des Installationsbefehls
92
Standard-Paket- oder Laufzeit-Installationspfad
Berechtigungsumfang
36
secrets or environment access, shell or command execution
Star-/Fork-Aktivität
57
163 Stars und 19 Forks; Issue-Aktivität ist in den aktuellen Metadaten nicht verfügbar
Akzeptanz
68
163 GitHub-Stars
Warnungen
- Dependency or permission surface needs review
- Permission surface may require sandboxing
- Quality score needs review
- Permission surface needs review: secrets or environment access, shell or command execution
- Stars/forks activity: 163 stars, 19 forks; issue activity unavailable in current metadata
- Dependency/runtime risk: command execution surface, credential or environment access
- Permission surface: secrets or environment access, shell or command execution
Methode
This report combines public metadata, AI review output, repository freshness, install readiness, OpenAgentSkill events, quality scoring, trust checks, and the agent safety gate. It is not a full source-code security review.
Nahe Optionen vergleichen
Ähnliche Skills als Nächstes prüfen
mono-color
Generate original one-ink or controlled two-ink editorial images from any theme, sentence, article idea, object, or reference photo. Always use this skill when the user asks for 单色海报、双色印刷、单色调视觉、蓝色/绿色孔版印刷、risograph、网点照片、复古或当代编辑排版、zine poster, monochrome editorial poster, duotone print, or asks to use the mono-color style. It uses an adaptive white, gray, or pale-beige substrate, no more than two printing inks, active negative space, terse human language, and strong serif/grotesk/mono typography without making retro styling the default or copying a source composition, wording, logo, or artwork. Produce both the final generation prompt and the generated raster image unless the user explicitly asks for prompt only.
1.9K Stars · Audit-Bericht
Last30days Skill
Research the last 30 days across Reddit, X, YouTube, Hacker News, Polymarket, GitHub, and the web, then synthesize a grounded brief for an AI agent.
64K Stars · Audit-Bericht
Academic Research Skills
Academic Research Skills for Claude Code: research → write → review → revise → finalize
38K Stars · Audit-Bericht