Laporan audit skill

ai-model-extraction Laporan audit.

Extract or steal an ML/LLM model's parameters, training data, or system prompt via query access — model stealing, membership inference, training-data extraction. Load when testing an ML API/endpoint, "model extraction/inversion", data-leakage or IP-theft concerns, exposed inference endpoints. Signals: a predict/inference API, embeddings endpoint, fine-tuned model.

Eksperimental · TinjauPerlu ditinjauDihasilkan 13 Okt 2026Audit metadata heuristik

Audit

Kepercayaan

Kualitas

Keamanan

100

Maintain

Pasang

Trust Score OpenAgentSkill

Tinjauan manual

Trust Score OpenAgentSkill

The Trust Score helps an agent decide whether a skill is safe enough to shortlist before installation.

Adopsi GitHub

Gagal

20 star GitHub

Aktivitas star/fork

Gagal

20 star dan 7 fork; aktivitas issue tidak tersedia dalam metadata saat ini

Pemeliharaan terbaru

Lulus

100

11 hari sejak push

Kejelasan lisensi

Lulus

MIT

Kelengkapan README/SKILL.md

Info

Metadata publik memerlukan konteks README/SKILL.md yang lebih kuat

Risiko dependensi/runtime

Info

credential or environment access, network or browser surface

Ketersediaan pemasangan

Lulus

npx skills add NoorQureshi/SploitAgent --skill ai-model-extraction

Keamanan perintah pemasangan

Lulus

Jalur pemasangan paket atau runtime standar

Cakupan izin

Peringatan

secrets or environment access, network or browser access

Bukti repositori

Lulus

https://github.com/NoorQureshi/SploitAgent/tree/main/skills/ai-ml/ai-model-extraction

Status peninjauan

Peringatan

Persetujuan tinjauan AI belum ada

Hasil terbukti Agent

Info

Belum ada data hasil Agent

Pemeriksaan

Tinjauan pemasangan dan adopsi

6 Lulus · 14 Perlu ditinjau

Jalur pemasangan

Lulus

npx skills add NoorQureshi/SploitAgent --skill ai-model-extraction

Repositori

Lulus

https://github.com/NoorQureshi/SploitAgent/tree/main/skills/ai-ml/ai-model-extraction

Lisensi

Lulus

MIT

Pemeliharaan

100

Lulus

11 hari sejak push

Tinjauan AI

Periksa

Review approval is missing

Kelengkapan README/SKILL.md

Lulus

Usable description available

Risiko dependensi

Periksa

credential or environment access, network or browser surface

Keamanan perintah pemasangan

Lulus

Jalur pemasangan paket atau runtime standar

Cakupan izin

Periksa

secrets or environment access, network or browser access

Aktivitas star/fork

Perbaiki

20 star dan 7 fork; aktivitas issue tidak tersedia dalam metadata saat ini

Adopsi

Periksa

20 star GitHub

Peringatan

Permission surface may require sandboxing
Low GitHub adoption signal
Persetujuan tinjauan AI belum ada
Quality score needs review
Permission surface needs review: secrets or environment access, network or browser access
GitHub adoption: 20 GitHub stars
Stars/forks activity: 20 stars, 7 forks; issue activity unavailable in current metadata
Permission surface: secrets or environment access, network or browser access
Review status: AI review approval is missing

Metode

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.

Jalur pemasangan

Tinjau laporan sebelum memasang pada Agent produksi.

$ npx skills add NoorQureshi/SploitAgent --skill ai-model-extraction

Metadata yang dapat dibaca Agent

Paket keputusan mesin

JSON tersemat menyediakan tugas dan Agent yang sesuai, perintah pemasangan, skor kepercayaan, gerbang keamanan, alternatif, serta kondisi jangan gunakan untuk ditinjau Agent.

Kebijakan pemasangan: Tinjau
Agent terbaik: Codex + Claude Code
Alternatif: 3

Buka JSON Agent

Sinyal

Skor audit: 73/100
Quality: 54/100
Trust: 69/100
Install safety: 92/100
Cakupan izin: 60/100
Pemeliharaan: 11 hari sejak push
Event: 10 event
Star GitHub: 20
Push terakhir: 2 Okt 2026

Keamanan Agent v2

45/100 · Hindari pemasangan otomatis

EksperimentalTinjau

Sparse or mixed signals. Useful for discovery, but not for autonomous installation.

Test manually in an isolated workspace and compare against safer alternatives.

Petunjuk izin berisiko tinggi: Secrets or environment access
skor keamanan Agent 45/100

Akses jaringan

Sedang

Skill kemungkinan mengambil halaman jarak jauh, API, repositori, atau layanan eksternal.

Secrets or environment access

Tinggi

Skill metadata references credentials, tokens, environment variables, or secret-bearing workflows.

Akses database

Sedang

Skill dapat memeriksa skema, mengkueri database, atau bekerja dengan penyimpanan persisten.

Petunjuk izin berisiko tinggi: Secrets or environment access
Permission surface may require sandboxing

Kembali ke skill Alternatif Badge audit SVG

Bandingkan opsi sekitar

Skill terkait untuk diaudit berikutnya

peft-fine-tuning

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

13K Star · Laporan audit

distributed-llm-pretraining-torchtitan

Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.

13K Star · Laporan audit

litert-lm

Creates an Android app that runs an open text LLM on the device, on the CPU or the GPU, with the LiteRT-LM Kotlin API. Use this skill to build a new chat, summarization or extraction app with a .litertlm model from Hugging Face litert-community (Gemma, Qwen, Llama, Phi) - the dependency and manifest entries, getting the model file onto the device, engine initialization, a streamed multi-turn conversation, the ViewModel and screen.

459 Star · Laporan audit