The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute jq, wc, ls, and sed for processing and aggregating data from newline-delimited JSON files. While these are standard tools, they are used to process various local system files.
[DATA_EXFILTRATION]: The skill accesses sensitive local file paths, including ~/.claude/history.jsonl (which contains command history) and session logs located in ~/.claude/projects/. These files can contain sensitive information about the user's activities, environment, and conversation history. Although the data is processed locally, this constitutes exposure of sensitive information.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes session logs (~/.claude/projects/**/*.jsonl) containing untrusted data from previous user and assistant messages.
Ingestion points: CC session logs and OrchestKit analytics files (e.g., agent-usage.jsonl, skill-usage.jsonl).
Boundary markers: None identified; there are no specific delimiters used to prevent the agent from interpreting content within session logs as instructions.
Capability inventory: Bash (used for jq, ls, wc, sed), Read, Grep, Glob.
Sanitization: None; the skill directly processes the log content using jq and presents the results.

analytics