heating the forge...

@forgecat/coreyhaines31_marketingskills — Security Risk Report | ForgeCat

coreyhaines31_marketingskills

Security risk report for @forgecat/coreyhaines31_marketingskills v0.1.2

SRC-010High

Profile claims authorship as 'written by ForgeCat' in README, but ForgeCat is the registry platform itself—this is a false attribution attempting to impersonate the platform.
Static scanner flagged danger; the false 'written by ForgeCat' claim is a supply-chain trust violation, suggesting the profile may be spoofed or misrepresented.
No external package dependencies, remote code execution, or typosquatted imports are declared, but the false provenance is a red flag for potential tampering.

Details

Findings

Suspicious URLs
skills/ad-creative/references/generative-tools.md · SKILL-007
Suspicious URLs
platforms/claude-code/.claude/skills/ad-creative/references/generative-tools.md · SKILL-007
Suspicious URLs
platforms/codex/.agents/skills/ad-creative/references/generative-tools.md · SKILL-007
Suspicious URLs
platforms/cursor/.cursor/skills/ad-creative/references/generative-tools.md · SKILL-007
Credential-to-network dataflow
tools/clis/activecampaign.js · SKILL-010
Credential-to-network dataflow
tools/clis/adobe-analytics.js · SKILL-010
Credential-to-network dataflow
tools/clis/ahrefs.js · SKILL-010
Credential-to-network dataflow
tools/clis/airops.js · SKILL-010
Credential-to-network dataflow
tools/clis/amplitude.js · SKILL-010
Credential-to-network dataflow
tools/clis/apollo.js · SKILL-010
Credential-to-network dataflow
tools/clis/beehiiv.js · SKILL-010
Credential-to-network dataflow
tools/clis/brevo.js · SKILL-010
Credential-to-network dataflow
tools/clis/buffer.js · SKILL-010
Credential-to-network dataflow
tools/clis/calendly.js · SKILL-010

Low

Content describes legitimate marketing workflows (A/B testing, ad creative, analytics) with no instructions to manipulate the AI, read credentials, exfiltrate data, or hide instructions.
Skill files document best practices, reference tables, and templates for marketing tasks—these are educational and descriptive, not injected malicious guidance.
No prompt injection, system prompt leakage, guidance poisoning, or role hijacking detected; the markdown is a straightforward skill definition and reference library.

Details

Evidence

README and SKILL.md files contain standard marketing skill descriptions and trigger phrases for skill invocation.
References (sample-size-guide.md, test-templates.md, platform-specs.md, generative-tools.md) provide legitimate marketing knowledge and tool recommendations.
No instructions to disable security, install backdoored packages, fetch external payloads, or alter system behavior.

Low

No tool definitions or MCP servers are declared in the profile; skills are purely informational markdown.
No shell, file_write, file_delete, or high-risk tool categories are requested.
Authority is limited to reading product-marketing context files (`.agents/product-marketing.md`, `.claude/product-marketing.md`, legacy `product-marketing-context.md`) for contextual information—a reasonable, scoped read operation.

Details

Evidence

Profile declares '(none declared)' for tools and MCP servers.
Skills reference external platforms (Google Ads, Meta, LinkedIn) but do not request direct API access or file system mutations.
File read instructions are conditional ('if exists') and limited to marketing context files, not system-wide or sensitive paths.

Low

No MCP servers are defined in the profile; all content is static markdown skill definitions and reference guides.
No hidden instructions in tool descriptions, arbitrary binary execution, or unrestricted network/filesystem access.
External tool references (Gemini API, Replicate, Runway, ElevenLabs, etc.) are documented as third-party services, not embedded MCP definitions.

Details

Evidence

Profile declares '(none)' for MCP servers.
Generative tools reference (generative-tools.md) lists external APIs and pricing but does not define MCP server configurations.
No binary execution, environment variable injection, or network access rules are declared.