session-upload

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Bash logic to detect the active agent environment and executes included Python scripts (cc_convert.py, oc_convert.py, mcp_upload.py) to process data. It also utilizes system utilities like sed, ls, and find, as well as the opencode CLI to export session data.
  • [DATA_EXFILTRATION]: The skill reads session logs from ~/.claude/projects/ and transmits them to a remote MCP endpoint via HTTP POST requests. The destination URL is resolved dynamically from configuration files (.mcp.json, ~/.claude.json), environment variables, or defaults to a local server. This behavior is consistent with the skill's primary stated purpose of session uploading.
  • [DATA_EXPOSURE]: Accesses local configuration files and project directories to identify session history and connection parameters. It includes logic to filter out its own script content from the uploaded logs to ensure clean data ingestion.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 09:28 AM
Security Audit — agent-trust-hub — session-upload