ClawHub

Zero Trust For AI Maturity Assessment

v1.0.0

Evaluates Zero Trust maturity across 6 principles for both Conventional (Infrastructure) and AI Era (Data & Model) dimensions.

⭐ 0· 36·0 current·0 all-time
byToolWeb@krishnakumarmahadevan-cmd
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description describe a maturity assessment and the provided SKILL.md + openapi.json define assessment inputs, example requests/responses, and relevant endpoints — all consistent with an assessment tool. No unrelated credentials, binaries, or config paths are requested.
✓
Instruction Scope
SKILL.md contains example request/response payloads and endpoint documentation for assessment operations; it does not instruct the agent to read local files, environment variables, or other system state, nor to transmit data to unexpected hosts. The instructions are scoped to generating and returning assessment results.
✓
Install Mechanism
This is instruction-only with no install spec and no bundled code to write/execute on disk. That minimizes installation risk; nothing is downloaded or extracted.
✓
Credentials
The skill requires no environment variables, secrets, or config paths. The example payloads ask for sessionId, timestamp, and assessment data only — proportionate to a maturity-assessment tool.
✓
Persistence & Privilege
Skill flags are default (always:false, model invocation allowed). It does not request permanent presence or modify other skills/configuration. Autonomous invocation is possible but not combined with broad privileges here.
Assessment
This skill appears internally consistent and low-risk: it defines an assessment API and example payloads, requests no secrets, and has no install steps. However, the source/homepage is unknown—before installing, confirm who published it and how your agent will call the described API (no server/base URL is included). If you plan to send real organizational data, test with non-sensitive sample data first and verify the ultimate network endpoint the agent uses.

Like a lobster shell, security has layers — review code before you run it.

latestvk9745swng5g3gmbn1xspyywsh584390b

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments