The Agent Skills Directory

[SAFE]: All 25 Python analysis tools provided in the suite (e.g., burn_rate_calculator.py, risk_quantifier.py, tech_debt_analyzer.py) are implemented using only Python's standard libraries. They perform deterministic calculations based on local user data and do not contain any network operations, external requests, or unauthorized file access.
[SAFE]: The suite implements a robust 'Agent Protocol' that defines explicit invocation syntax, depth limits (max 2), and circular call prevention. This ensures that the AI agent team can coordinate cross-functional analysis safely without triggering infinite loops or cascading errors.
[SAFE]: The skill includes a detailed 'Anonymization Protocol' designed to strip sensitive company information (such as exact revenue figures, customer names, and runway months) before any data is sent to external tools or web searches, demonstrating a privacy-first architecture.
[SAFE]: Data persistence is handled locally through structured markdown files in the user's home directory (e.g., ~/.claude/company-context.md and ~/.claude/decision-log.md). These files are used to maintain context across sessions without requiring external database connections.
[SAFE]: The 'Internal Quality Loop' enforces self-verification, peer-verification, and critic pre-screening of all outputs before they reach the user, reducing the risk of hallucinated or inconsistent advice.
[SAFE]: No evidence of prompt injection, obfuscation, or unauthorized privilege escalation was found across the 124 files analyzed. All scripts and instructions are consistent with the stated purpose of executive leadership advisory.

c-level-advisor