The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a bundled Python script (scripts/humanizer_scorer.py) to calculate a humanity score for text drafts. Technical review of the script confirms it uses only standard libraries for regex and math and performs no network operations or unsafe system calls.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user-provided drafts and a brand context file (.claude/product-marketing-context.md). 1. Ingestion points: User input for 'the content' and the local marketing context file. 2. Boundary markers: Absent for the context file; instructions suggest pasting the draft directly into the session. 3. Capability inventory: The agent can execute a local Python script for text analysis. 4. Sanitization: No explicit validation or sanitization is performed on the ingested content before processing.
[DATA_EXPOSURE]: The skill accesses a sensitive environment path (.claude/product-marketing-context.md) to retrieve marketing guidelines. While this file contains brand-specific data, the access is localized to the agent's environment and matches the stated purpose of the skill.

content-humanizer