lode-intent-sync
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/intent_targets.py) to discover relevant intent artifacts within the repository. It also references a shared utility (scripts/lode_raw.py) for managing project metadata if present in the environment. - [DATA_EXFILTRATION]: Accesses repository documentation (e.g.,
DESIGN.md,PLAN.md) and configuration files in~/.lode/or the project root. This data is used locally to generate summaries and proposals for the user; no external network transmission or data exfiltration was detected. - [PROMPT_INJECTION]: The skill ingests and processes markdown content from the repository, which constitutes an untrusted data source and creates a surface for indirect prompt injection. However, the skill mitigates this risk by enforcing a mandatory human-in-the-loop review, requiring explicit user approval for all proposed file writes.
Audit Metadata