OpenAgentSkill
GitHub200+

Evaluasi sebelum pemasangan

Visual Compile Laporan evaluasi.

Keputusan pemasangan yang dapat dibaca mesin untuk Agent: kecocokan tugas, Trust Score, Audit Score, keamanan pemasangan, cakupan izin, dan rencana validasi konkret sebelum skill menyentuh workspace.

Perlu ditinjauRisiko sedangTinjau Kebijakan
73
Laporan evaluasi
76
Kepercayaan
80
Audit
64
Keamanan Agent

Tinjauan manual

Memerlukan persetujuan manusia sebelum memasang ke workspace nyata.

Gerbang wajib

Pemeriksaan yang harus dilalui Agent sebelum memasang

Buka JSON

Kecocokan tugas

70

Peringatan

Task fit is weak; compare alternatives before selecting.

  • Evaluate Visual Compile before installing it in an agent workflow
  • Desain dan kreatif
  • Design and creative workflows; Claude Code teams; teams that value GitHub adoption signals

Jalur pemasangan

92

Lulus

Install handoff is available.

  • npx skills add Player-YN/PawWork_ZhuaZhua --skill Visual Compile

Keamanan perintah pemasangan

92

Lulus

Jalur pemasangan paket atau runtime standar

  • npx skills add Player-YN/PawWork_ZhuaZhua --skill Visual Compile

Skor kepercayaan

76

Peringatan

Good trust signals with a few areas worth checking before rollout.

  • Shortlist kuat
  • 1.3K star GitHub
  • MIT

Skor audit

80

Peringatan

Perlu ditinjau

  • The skill is tightly coupled to a specific host platform (Paw Work) and its internal tools (inspect, fromRaster, deck, etc.), which limits portability to other agent frameworks. However, this is acceptable for a platform-specific skill.

Gerbang keamanan Agent

64

Peringatan

Kandidat yang dapat digunakan, tetapi Agent harus menampilkan catatan izin dan audit sebelum memasang.

  • Memerlukan persetujuan manusia sebelum memasang ke workspace nyata.
  • The skill is tightly coupled to a specific host platform (Paw Work) and its internal tools (inspect, fromRaster, deck, etc.), which limits portability to other agent frameworks. However, this is acceptable for a platform-specific skill.

Kejelasan lisensi

86

Lulus

MIT

  • MIT

Cakupan izin

72

Peringatan

filesystem or document access, network or browser access

  • Network access: medium
  • Filesystem access: medium

Rencana validasi

Langkah Agent berikutnya

  1. 1Inspect repository, README/SKILL.md, license, and recent commits before production use.
  2. 2Install in an isolated workspace or sandbox with no production secrets available.
  3. 3Run the smallest representative task and record files touched, commands run, network access, and outputs.
  4. 4Compare the selected skill against at least one alternative when the eval status is review or failed.
  5. 5Promote only after the agent reports a successful verification result and unresolved warnings are accepted.

Jangan gunakan ketika

Kondisi yang membutuhkan skill lain

  • Tim yang membutuhkan SLA dengan dukungan vendor
  • production agents without a repository review
  • The skill is tightly coupled to a specific host platform (Paw Work) and its internal tools (inspect, fromRaster, deck, etc.), which limits portability to other agent frameworks. However, this is acceptable for a platform-specific skill.
  • Quality score needs review
  • Production credentials, payments, or irreversible account changes without explicit human review
  • Sensitive private data before reviewing repository code, license, and permission surface

Pemeriksaan pendukung

Sinyal kepercayaan di balik keputusan

Kelengkapan README/SKILL.md

Peringatan

76

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

Pemeliharaan terbaru

Lulus

88

1 bulan sejak push

Alternatif tersedia

Lulus

82

Alternative skills are available for comparison.