51mee Background Check List
v1.2.0背景调查清单。触发场景：用户准备录用候选人，要求生成背调清单和背调问题模板。
⭐ 0· 127·0 current·0 all-time
by51mee@51mee-com
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name/description match the SKILL.md: it parses a resume and produces a background-check checklist, questions, risk flags, and a report template. There are no unrelated binaries, environment variables, or config paths requested.
Instruction Scope
Runtime instructions are narrowly scoped to parsing resume text and producing structured JSON/markdown output. They do not reference system files, secrets, network endpoints, or external upload targets. The SKILL.md claims local-only model execution and automatic data clearing — reasonable as design goals but these are promises in documentation rather than enforceable actions by an instruction-only skill.
Install Mechanism
No install spec and no code files are present, so nothing will be written to disk or downloaded during installation.
Credentials
The skill requests no environment variables, credentials, or config paths. This matches its simple processing purpose.
Persistence & Privilege
always is false and the skill does not request elevated persistence or to modify other skills/config. Autonomous invocation is allowed (platform default) but not combined with additional privileges.
Assessment
This skill appears coherent and low-risk in structure, but note that privacy claims in SKILL.md (local-only inference, automatic clearing, never saving personal data) are statements by the author and not enforced by the skill itself. Before using with real candidate data: (1) confirm your OpenClaw/agent is configured for on‑device/local model inference if you require that guarantee; (2) obtain candidate consent and follow applicable privacy laws; (3) test the skill with non-sensitive example resumes to confirm output format and that it doesn't hallucinate details; and (4) if you need auditability, verify where logs and outputs are stored by your agent/platform. If any additional network, install, or credential requests appear later, treat that as a suspicious change.Like a lobster shell, security has layers — review code before you run it.
latestvk97f1z52sxvs852jr36as903x983201a
License
MIT-0
Free to use, modify, and redistribute. No attribution required.