The Agent Skills Directory

[SAFE]: The skill's primary function is to provide educational content and a local utility script for project management. No malicious patterns were identified across the provided documentation and scripts.
[COMMAND_EXECUTION]: The Python script scripts/refinement_scorer.py is designed to be executed by the user. A detailed review of the source code confirms it uses standard libraries (json, argparse, re, datetime) to perform heuristic analysis on local JSON files. It does not invoke subshells, execute arbitrary commands, or perform unsafe deserialization.
[DATA_EXFILTRATION]: The skill does not contain any network-communication code, such as requests, urllib, or socket calls. It also does not include hardcoded credentials or access sensitive system paths.
[PROMPT_INJECTION]: The instructions in SKILL.md and the references focus entirely on agile methodologies. There are no attempts to override agent behavior, bypass safety filters, or extract system prompts.
[EXTERNAL_DOWNLOADS]: The skill does not perform any remote package installations or download external scripts during runtime.

backlog-refinement