Security risk report for @forgecat/openai_skills_system-skill-creator v0.0.11
Source Integrity
Low
Profile claims OpenAI as author with a GitHub repository link; metadata is consistent and verifiable.
No suspicious or obfuscated source indicators; standard skill documentation structure.
Agent Intent
Low
Content is educational guidance on skill creation best practices—no instructions to manipulate the AI, exfiltrate data, or bypass security controls.
Describes skill architecture, context management, and design principles without injecting malicious directives or poisoned knowledge.
No prompt injection, system prompt leakage, or guidance poisoning detected; the skill teaches legitimate skill-building methodology.
Details
Evidence
Core content focuses on 'Concise is Key', 'Set Appropriate Degrees of Freedom', and 'Progressive Disclosure Design Principle'—all standard software engineering practices.
References to 'SKILL.md', 'scripts/', 'references/', and 'assets/' are structural guidance, not hidden instructions.
No directives to read credentials, fetch external payloads, or weaken security defaults.
Permissions
Low
No tools or MCP servers declared; the profile is purely instructional documentation.
No file_write, file_delete, shell, or web_fetch permissions requested or implied.
Authority scope matches stated purpose: guiding users to create or update skills.
Details
Evidence
Declared tools section is empty.
MCP servers section is empty.
Content describes how to structure skills, not how to execute privileged operations.
MCP Risk
Low
No MCP servers defined in the profile.
No binary execution, network access, or filesystem operations configured.
Profile is a static documentation skill with no runtime dependencies.
Details
Evidence
MCP servers section is empty.
Dependencies listed as 'None' in the README metadata.
ForgeCat