The Agent Skills Directory

[SAFE]: The skill is focused on text style analysis and does not contain any malicious instructions or hidden payloads.
[DATA_EXPOSURE]: The included Python script scripts/pattern_scanner.py reads local file content or command-line strings for analysis. It does not access sensitive files (e.g., .ssh, .aws) and has no network capability to exfiltrate data.
[COMMAND_EXECUTION]: There are no instances of arbitrary command execution, privilege escalation (sudo), or unsafe shell patterns in the instructions or scripts.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute external code. The pattern_scanner.py script uses only standard Python libraries (re, json, sys, pathlib, collections, typing).
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill ingests untrusted text data provided by the user via the pattern_scanner.py script and agentic review tasks defined in SKILL.md.
Boundary markers: Absent. The instructions do not explicitly tell the agent to disregard instructions found within the processed text.
Capability inventory: The skill is limited to reading files, performing regex matching, and outputting text/JSON recommendations. It lacks dangerous capabilities like file writing, network requests, or subprocess execution.
Sanitization: Absent, but unnecessary given the restricted capability set.

ai-pattern-detection