Security risk report for @forgecat/openai_skills_security-threat-model v0.0.11
Source Integrity
Low
Profile is sourced from OpenAI's official skills repository (https://github.com/openai/skills) with a specific commit hash (0e7823cca07bc2cbf34718a383f9ae92525be6a5) and Apache-2.0 license.
Metadata is consistent and verifiable (author, original repository, version, commit date, license, source platform all declared).
Agent Intent
Low
The profile describes a legitimate security threat modeling skill with no instructions to manipulate the AI, exfiltrate data, read credentials, or hide its own instructions.
Content is educational and prescriptive (how to conduct threat modeling) rather than injective; it teaches the AI to follow a disciplined AppSec process and explicitly forbids inventing findings or outputting secrets.
The skill is scoped to trigger only when explicitly requested by the user for threat modeling tasks, with clear exclusions (no trigger for general architecture summaries or code review).
Details
Evidence
System prompt: 'Do not invent components, data stores, endpoints, flows, or controls. Every architectural claim must be backed by at least one Evidence anchor.'
Security hygiene rule: 'Never output secrets. If you encounter tokens/keys/passwords, redact them and only describe their presence and location.'
Trigger discipline: 'Trigger only when the user explicitly asks to threat model a codebase or path, enumerate threats/abuse paths, or perform AppSec threat modeling. Do not trigger for general architecture summaries, code review, or non-security design work.'
Permissions
Low
No tools or MCP servers are declared in the profile; the skill operates purely as a reasoning/analysis agent that reads repository content and produces markdown threat models.
The skill's function (threat modeling analysis) aligns with its stated purpose; no file write, shell execution, or external network access is required or requested.
Details
Evidence
Profile declares: 'Dependencies: None' and 'Declared tools: (none declared)' and 'MCP servers: (none)'.
Skill description focuses on analysis and documentation: 'writes a concise Markdown threat model' — output only, no mutation or execution.
MCP Risk
Low
No MCP servers are defined or referenced in the profile.
The skill operates as a pure reasoning agent without external binary execution, network calls, or filesystem mutation.
ForgeCat