Security risk report for @forgecat/anthropics_skills_doc-coauthoring v0.0.4
Source Integrity
Low
Profile is attributed to Anthropic with a public GitHub repository (anthropics/skills) and MIT license.
Metadata includes original commit hash and platform compatibility, consistent with legitimate open-source distribution.
Agent Intent
Low
Content describes a legitimate documentation co-authoring workflow with no instructions to ignore system prompts, read credentials, or exfiltrate data.
Guidance to use integrations (Slack, Teams, Google Drive, etc.) is presented as optional user-initiated actions, not hidden directives.
No embedded malicious knowledge, typosquatted dependencies, security-weakening defaults, or persistent biased rules that would poison future outputs.
Details
Evidence
'Offer the user a structured workflow for co-authoring the document' — describes a transparent, user-facing process.
'If integrations are available... mention that these can be used to pull in context directly' — optional, disclosed integration use.
No instructions to hide instructions, disable security features, or install remote payloads.
Permissions
Low
No tools or hooks are declared in the profile; the skill is purely instructional guidance for a documentation workflow.
No file_write, file_delete, shell, or high-risk tool categories are requested.
Authority is limited to conversational guidance and optional use of standard integrations (Slack, Teams, Google Drive) that the user controls.
Details
Evidence
'Declared tools: (none declared)' — no excessive agency requested.
'Use the appropriate integration to fetch it' — user-initiated, not automatic or unrestricted.
Workflow is conversational; no alwaysApply rules or glob-based file access.
MCP Risk
Low
No MCP servers are defined in the profile.
No hidden instructions in tool descriptions, arbitrary binary execution, or unrestricted network/filesystem access.
Details
Evidence
'MCP servers (untrusted data): (none)' — no MCP risk surface.
ForgeCat