The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the find command to identify and execute a shell script (half-clone-conversation.sh) within the ~/.claude directory. Executing scripts from the filesystem via dynamic path resolution is a high-capability operation.
[DATA_EXFILTRATION]: The skill reads from ~/.claude/history.jsonl to extract session IDs and project paths. While the data is processed locally, conversation history is considered sensitive, and accessing it constitutes a data exposure risk.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes data from external history files that may contain instructions.
Ingestion points: ~/.claude/history.jsonl (referenced in SKILL.md)
Boundary markers: None provided; the skill relies on tail and jq for parsing.
Capability inventory: Includes execution of shell utilities (tail, jq, find) and the execution of a bash script.
Sanitization: There is no evidence of sanitization or validation for the metadata extracted from the history file before it is used as command arguments.

quarter-clone