book-metrics-generator
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes multiple user-controlled files to generate metrics, creating a surface for indirect prompt injection.
- Ingestion points: Reads content from
docs/chapters/**/*.md,docs/learning-graph/learning-graph.csv,docs/glossary.md,docs/faq.md,docs/sims/**/index.md, anddocs/chapters/**/quiz.md. - Boundary markers: The script does not implement specific delimiters or safety warnings when reading these files.
- Capability inventory: The skill executes a shell script (
book-metrics-generator.sh), a Python script (book-metrics.py), and writes two markdown files (book-metrics.mdandchapter-metrics.md) to the local file system. - Sanitization: While the word counting logic filters code blocks and URLs, the chapter title extraction (
_extract_title) and other header-based counts use raw regex matches on file content without sanitization or escaping. - [COMMAND_EXECUTION]: The skill relies on the execution of a bash wrapper script and a Python script to perform its analysis. The
SKILL.mdfile explicitly instructs the user on how to run these scripts directly from the terminal, which is a standard part of its documented functionality.
Audit Metadata