Skill-Audit-Bericht
accuracy-safe-quantization Audit-Bericht.
Shrink a converted LiteRT model with ai-edge-quantizer (fp16 / int8 / int4) without losing accuracy, verifying parity against the float source after every step. Use when choosing a quantization recipe for a new model, when a quantized model fails to load, degrades on a task benchmark, or degenerates over long generations, or when deciding between dynamic-range, weight-only, and blockwise variants.
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
459 GitHub-Stars
Star-/Fork-Aktivität
Info62
459 Stars und 122 Forks; Issue-Aktivität ist in den aktuellen Metadaten nicht verfügbar
Aktuelle Wartung
Bestanden100
2 Tage seit dem letzten Push
Lizenzklarheit
Bestanden86
Apache-2.0
README/SKILL.md-Vollständigkeit
Bestanden86
Metadaten enthalten ausreichend Nutzungs- und Workflow-Kontext
Abhängigkeits-/Laufzeitrisiko
Fehlgeschlagen36
command execution surface, credential or environment access
Installationsverfügbarkeit
Bestanden92
npx skills add google-ai-edge/litert-samples --skill accuracy-safe-quantization
Sicherheit des Installationsbefehls
Bestanden92
Standard-Paket- oder Laufzeit-Installationspfad
Berechtigungsumfang
Fehlgeschlagen22
secrets or environment access, shell or command execution
Repository-Nachweis
Bestanden86
https://github.com/google-ai-edge/litert-samples/tree/main/skills/accuracy-safe-quantization
Prüfstatus
Warnung46
KI-Prüffreigabe fehlt
Agent-validierte Ergebnisse
Info54
Noch keine Agent-Ergebnisdaten
Prüfungen
Installations- und Adoptionsprüfung
Installationspfad
92
npx skills add google-ai-edge/litert-samples --skill accuracy-safe-quantization
Repository
88
https://github.com/google-ai-edge/litert-samples/tree/main/skills/accuracy-safe-quantization
Lizenz
86
Apache-2.0
Wartung
100
2 Tage seit dem letzten Push
KI-Prüfung
55
Review approval is missing
README/SKILL.md-Vollständigkeit
86
Usable description available
Abhängigkeitsrisiko
36
command execution surface, credential or environment access
Sicherheit des Installationsbefehls
92
Standard-Paket- oder Laufzeit-Installationspfad
Berechtigungsumfang
22
secrets or environment access, shell or command execution
Star-/Fork-Aktivität
62
459 Stars und 122 Forks; Issue-Aktivität ist in den aktuellen Metadaten nicht verfügbar
Akzeptanz
68
459 GitHub-Stars
Warnungen
- Dependency or permission surface needs review
- Permission surface may require sandboxing
- Potential broker, wallet, exchange, or real-money execution surface; sandbox and explicit approval are required
- KI-Prüffreigabe fehlt
- This skill may touch real-money trading, broker, wallet, or exchange operations; use only in a sandbox with explicit approval.
- Quality score needs review
- Permission surface needs review: secrets or environment access, shell or command execution
- Dependency/runtime risk: command execution surface, credential or environment access
- Permission surface: secrets or environment access, shell or command execution
- Review status: AI review approval is missing
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
weather
Current weather and forecasts with web_fetch, falling back to wttr.in curl for locations, rain, temperature, travel planning.
391K Stars · Audit-Bericht
mcporter
List, configure, authenticate, call, and inspect MCP servers/tools with mcporter over HTTP or stdio.
391K Stars · Audit-Bericht
release-openspec
Use this skill when releasing OpenSpec: audit merged work and changeset coverage, decide whether a catch-up changeset PR is needed, prepare or resume the Changesets Version Packages PR, cut a beta or stable release, verify publishing, and polish GitHub release notes. Also use when asked whether an open release PR is complete, what the next release step is, or to continue a release paused for human approval.
71K Stars · Audit-Bericht