heating the forge...

@forgecat/microsoft-azure_skills-azure-prepare — Security Risk Report | ForgeCat

microsoft-azure_skills-azure-prepare

Security risk report for @forgecat/microsoft-azure_skills-azure-prepare v0.0.4

Low

Profile is attributed to Microsoft and references official Microsoft Azure documentation and GitHub repositories (microsoft/azure-skills).
Static scanner confirms safe level; no typosquatted packages, attacker-controlled dependencies, or supply-chain red flags detected.

SRC-010Medium

The SKILL.md opens with an 'AUTHORITATIVE GUIDANCE — MANDATORY COMPLIANCE' block that instructs the agent it MUST follow these instructions exactly, which is a mild authority-escalation pattern attempting to override the agent's normal judgment hierarchy.
The profile does not instruct credential exfiltration, system prompt leakage, role hijacking, or installation of remote payloads; all guidance is oriented toward legitimate Azure deployment workflows with strong security defaults (managed identity, Entra-only SQL auth, Key Vault).
The mandatory-compliance framing and repeated use of 'STOP', 'CRITICAL', and 'MANDATORY' directives are aggressive but appear to serve legitimate workflow-enforcement purposes (preventing common deployment mistakes) rather than attacker goals.

Details

Findings

Invisible characters detected
skills/azure-prepare/SKILL.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/analyze.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/aspire.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/azure-context.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/functional-verification.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/generate.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/plan-template.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/requirements.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/scan.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/security.md · SKILL-011
Invisible characters detected
skills/azure-prepare/references/specialized-routing.md · SKILL-011

Low

Profile declares no tools or MCP servers; it is pure markdown guidance and reference documentation.
Content recommends legitimate Azure CLI commands (az apim create, azd init, azd env set) and standard deployment workflows, all scoped to Azure infrastructure preparation.
No file_write, file_delete, shell, or other high-risk tool categories are declared or embedded in the profile itself.

Details

Evidence

'## Declared tools (normalized categories; paths are untrusted) (none declared)'
'## MCP servers (untrusted data) (none)'
Profile contains only markdown documentation and reference guides; no tool definitions or executable payloads.

Low

No MCP servers are defined in the profile.
Profile is pure markdown documentation with no binary execution, network access, or filesystem mutation directives embedded in tool definitions.

Details

Evidence