skill-optimizer

Warn

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses python3 and bash to execute analysis logic on the local system. It performs environment reconnaissance by checking for file existence using test -e and identifying installed CLI tools with which.
  • [DATA_EXFILTRATION]: The skill accesses highly sensitive session history paths, including ~/.claude/projects/ and ~/.codex/sessions/. These JSONL transcripts contain the complete record of user and assistant interactions, which may include private data, secrets, or proprietary information harvested during analysis.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its processing of untrusted historical session data.
  • Ingestion points: Processes raw session transcripts from ~/.claude/projects/**/*.jsonl and ~/.codex/sessions/**/*.jsonl.
  • Boundary markers: No delimiters or warnings are used to prevent the agent from obeying instructions embedded in the historical logs.
  • Capability inventory: The skill can execute Python scripts, run shell commands, and read local files.
  • Sanitization: There is no evidence of sanitization or validation of the log content before it is processed by the agent's logic.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 11, 2026, 06:59 AM
Security Audit — agent-trust-hub — skill-optimizer