CDO / Chief Data Officer
v1.0.1Drive data strategy with governance frameworks, analytics platforms, AI/ML initiatives, and privacy compliance.
⭐ 2· 698·0 current·0 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name and description (data strategy, governance, analytics, ML, privacy) match the included SKILL.md and the five supporting files (strategy.md, governance.md, analytics.md, ml.md, privacy.md). All required content is proportional to a Chief Data Officer advisory skill; there are no unrelated requirements (no cloud creds, no unrelated binaries).
Instruction Scope
The runtime instructions ask the agent to act as a virtual CDO and point to topical guidance in the included files. The instructions do not direct the agent to read system files, environment variables, or external endpoints beyond optional use of the platform's 'clawhub install' when the user explicitly confirms. High‑risk actions (vendor selection, incident response, data monetization) are explicitly marked as human-in-the-loop.
Install Mechanism
There is no install spec and no code files; this instruction-only skill writes nothing to disk and does not download or execute third-party packages. That minimizes its install-time risk.
Credentials
The skill declares no required environment variables, no primary credential, and no config paths. SKILL.md does not attempt to access undeclared secrets. The lack of requested credentials is proportionate to a purely advisory skill.
Persistence & Privilege
always:false (default) and user-invocable:true. The skill does not request permanent presence or elevated platform privileges. Note: autonomous invocation (disable-model-invocation:false) is the platform default; this skill being able to be called autonomously is normal and not, by itself, a red flag.
Assessment
This skill is advisory-only and coherent with its purpose. Before installing or using it: (1) remember it does not request credentials, but any recommendations it gives may lead you to grant access to data systems — review and approve such access yourself; (2) treat vendor/tool recommendations as starting points, verify vendor trust and licensing before onboarding; (3) decline or carefully vet any follow-up prompts that ask you to paste secrets, run shell commands, or install third-party packages; (4) keep the human-in-the-loop items (incident response, vendor selection, data monetization) under direct human control rather than relying solely on the agent.Like a lobster shell, security has layers — review code before you run it.
latestvk97a9b7mmvq6y539rd5r7yxg7181h260
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
📊 Clawdis
OSLinux · macOS · Windows