Back to skill
Skillv1.0.0
ClawScan security
Internal Comms Anthropic · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 14, 2026, 3:37 PM
- Verdict
- Benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill is an instruction-only set of writing guidelines for internal communications and its requirements and files align with that purpose.
- Guidance
- This skill is a coherent, instruction-only template pack for internal communications. Before enabling it: 1) Verify what connectors (Slack, Google Drive, Email, Calendar) the agent will use — confirm which account(s) and scopes are granted so the skill cannot read more than necessary. 2) Consider requiring user confirmation before the agent retrieves documents or messages (to avoid unwanted data exposure). 3) Test the skill on non-sensitive/dummy content to confirm formatting and behavior. 4) Be mindful when drafting incident reports or other sensitive comms — ensure the agent isn't including confidential details from source documents without review. Finally, note the included LICENSE is Apache-2.0.
Review Dimensions
- Purpose & Capability
- okName/description match the included files: SKILL.md plus example templates for 3P updates, newsletters, FAQs and general comms. There are no unexpected environment variables, binaries, or install steps requested — everything present is consistent with a formatting/guideline skill.
- Instruction Scope
- noteRuntime instructions ask the agent to 'pull from available sources' (Slack, Google Drive, Email, Calendar) when possible. That is coherent for producing internal comms, but it broadens the agent's data-access surface: the skill expects access to company communication tools though it does not itself declare or manage authentication for them.
- Install Mechanism
- okNo install spec and no code files — instruction-only. This is the lowest-risk install profile (nothing is written to disk by the skill).
- Credentials
- noteThe skill requires no environment variables or credentials itself, which is proportionate. However, the instructions assume access to Slack/Drive/Email/Calendar; you should confirm what connector credentials the agent/platform will use and whether those grants are appropriately scoped.
- Persistence & Privilege
- okalways is false and the skill is user-invocable; model invocation is enabled (normal default). This is appropriate for an assistant capability. If the platform grants the skill autonomous access to company connectors, that increases impact but the skill metadata itself does not request persistent elevated privileges.