capture
Pass
Audited by Gen Agent Trust Hub on Jun 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is to summarize conversation context and store it locally. All file system operations are constrained to local configuration paths (~/.lode, ~/.weekly-ppt) and the user-specified knowledge vault. No evidence of data exfiltration or network activity was found.
- [COMMAND_EXECUTION]: The skill executes a local Python helper script (
scripts/lode_raw.py) provided within the skill package. This script is used for deterministic path resolution and JSON data storage. Analysis of the script confirms it uses safe parsing (json.loads and yaml.safe_load) and does not perform any dangerous operations. - [SAFE]: Instructions regarding the modification of
~/.claude/settings.jsonfor auto-capture hooks are documented and transparent, serving the stated purpose of the skill to automate session recording.
Audit Metadata