Communication DNA

• Risk level: Low
• Analyzed: Feb 25, 2026, 2:04 AM
• Findings: 0
• Version: 1.0.1

Overview

The skill's code and runtime instructions align with its stated purpose (transcript ingestion, analysis, speaker profiling and optional CRM/KB linking), but it reads/writes sibling local databases and can import/execute code from a nearby Knowledge Base project — review those integrations and run in an isolated environment before use.

This skill appears to do what it says (ingest transcripts, analyze speakers, and optionally link to a local CRM and Knowledge Base). Before installing or running it: - Inspect the included scripts (you have them) and pay special attention to scripts/integrations.py: it opens ../personal-crm/crm.db and ../knowledge-base/knowledge.db. If those files exist, the skill will read them; the push-to-kb path will import and execute code from the knowledge-base package (sys.path manipulation). Only enable CRM/KB linking if you trust those local projects. - Run the skill in an isolated environment (container or VM) if you will ingest sensitive transcripts. The web UI (app.py) binds to port 5053 by default — avoid exposing it publicly. - The skill will create communication_dna.db in scripts/ and an uploads/ directory under scripts/. Confirm you’re comfortable with these files being written in the workspace. - If you don’t want cross-system access, do not place crm.db or knowledge.db in the relative sibling paths or remove/rename the integrations, or run the skill with read access restricted to those directories. - If you need higher assurance, review the omitted/truncated files referenced in SKILL.md (they were not fully shown) and audit any code in your local knowledge-base/personal-crm projects that could be executed by this skill.