[COMMAND_EXECUTION]: The skill utilizes a local script scripts/search_corpus_examples.py for database lookups. This script is part of the skill's infrastructure and is used as intended to provide relevant writing examples to the user.
[INDIRECT_PROMPT_INJECTION]: The skill's architecture involves reading and analyzing external novel excerpts from the data/articles/ directory. While these files are local to the skill, they represent processed external data. Ingestion points: The agent reads various .txt files in data/articles/ and analysis/excerpts.csv. Boundary markers: No explicit delimiters are used in the prompt instructions to isolate external text from system logic. Capability inventory: The skill invokes local search scripts via shell commands. Sanitization: There is no mention of filtering or sanitizing the input data.
[SAFE]: No critical vulnerabilities or malicious patterns such as remote code execution, credential harvesting, or data exfiltration were found. The Record Numbers (备案号) found in the data files are legitimate platform identifiers and do not contain hidden payloads.

webnovel-writing