agency-dashboard
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts to aggregate campaign status, budget pacing, and team capacity. Evidence: In SKILL.md, the process invokes campaign-tracker.py, execution-tracker.py, and team-manager.py to retrieve operational data.
- [PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection through the ingestion of external brand data and guidelines. 1. Ingestion points: Data is read from ~/.claude-marketing/brands/{slug}/profile.json, guidelines/_manifest.json, and outputs from tracking scripts. 2. Boundary markers: Absent; no specific delimiters are used to isolate untrusted brand data from instructions. 3. Capability inventory: The skill can execute local scripts and read internal application files. 4. Sanitization: Absent; no validation or sanitization of ingested brand data is described in the process.
Audit Metadata