OpenAgentSkill
GitHub200+

Évaluation avant installation

csdn-technical-writing Rapport d’évaluation.

Une décision d’installation lisible par machine pour les Agents : adéquation à la tâche, Trust Score, Audit Score, sécurité d’installation, surface de permissions et plan de validation concret avant de toucher l’espace de travail.

Revue nécessaireRisque moyenRevoir Politique
69
Rapport d’évaluation
73
Confiance
77
Audit
57
Sécurité Agent

Revue manuelle

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

Gates requis

Vérifications qu’un Agent doit réussir avant installation

Ouvrir JSON

Adéquation à la tâche

84

Validé

Task wording matches this skill metadata.

  • Evaluate csdn-technical-writing before installing it in an agent workflow
  • Sécurité
  • Content automation workflows; Claude Code teams; builders willing to evaluate younger projects

Chemin d’installation

92

Validé

Install handoff is available.

  • npx skills add FAIRY123456789/human-edge-agent-skills --skill csdn-technical-writing

Sécurité de la commande d’installation

92

Validé

Chemin d’installation standard de package ou runtime

  • npx skills add FAIRY123456789/human-edge-agent-skills --skill csdn-technical-writing

Score de confiance

73

Avertissement

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

  • Liste solide
  • 27 stars GitHub
  • MIT

Score d’audit

77

Avertissement

Revue nécessaire

  • SKILL.md does not explicitly document the expected input/output contract for the assistant, such as whether the skill should create a new Markdown draft, rewrite an existing one, or only audit and report issues.

Contrôle de sécurité Agent

57

Avertissement

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

  • Test manually in an isolated workspace and compare against safer alternatives.
  • SKILL.md does not explicitly document the expected input/output contract for the assistant, such as whether the skill should create a new Markdown draft, rewrite an existing one, or only audit and report issues.

Clarté de licence

86

Validé

MIT

  • MIT

Surface de permissions

100

Validé

Aucune surface de permissions à haut risque dans les métadonnées publiques

  • Browser automation: medium
  • Network access: medium
  • Filesystem access: medium

Plan de validation

Ce que l’Agent doit faire ensuite

  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.

Ne pas utiliser quand

Conditions exigeant une autre skill

  • Équipes qui nécessitent un SLA soutenu par le fournisseur
  • production agents without a repository review
  • Low GitHub adoption signal
  • SKILL.md does not explicitly document the expected input/output contract for the assistant, such as whether the skill should create a new Markdown draft, rewrite an existing one, or only audit and report issues.
  • The lint script is heuristic and only checks structural formatting; SKILL.md does not state that the script is advisory and does not validate the teaching sequence or semantic quality.
  • Quality score needs review

Vérifications complémentaires

Signaux de confiance derrière la décision

Complétude README/SKILL.md

Validé

86

Les métadonnées incluent suffisamment de contexte d’usage et de workflow

Maintenance récente

Validé

100

29 jours depuis le dernier push

Alternatives disponibles

Validé

82

Alternative skills are available for comparison.