review-writer
v1.0.1Data-driven writing and polishing for professional workplace documents. Use when the user asks to write or refine: (1) 述职/复盘/年终总结, (2) 简历/项目经历/STAR 改写, (3) 周...
⭐ 0· 22·0 current·0 all-time
bystrive@ykitty
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name, description, and templates all describe workplace document drafting (述职、简历、周报、OKR、PRD etc.). The skill declares no binaries, env vars, or config paths—this is proportionate and expected for a writing helper.
Instruction Scope
SKILL.md provides detailed, bounded runtime instructions (produce a 70% first draft by default, ask up to ~5 targeted questions when needed, use specific templates and a data-driven checklist). It does not instruct reading system files, environment variables, or sending data to external endpoints. Minor functional risk: the default behavior of producing a draft even when some metrics are missing can lead to placeholders or fabricated numbers unless the user requires clarifying questions first—this is a correctness/privacy concern rather than a security incoherence.
Install Mechanism
Instruction-only skill with no install spec and no code files. Nothing is written to disk; this is the lowest-risk install model.
Credentials
The skill requests no environment variables, credentials, or config paths. There is no disproportionate access requested for its stated purpose.
Persistence & Privilege
always is false and model invocation is allowed (platform default). The skill does not request permanent presence, nor does it instruct modifying other skills or system-wide settings.
Assessment
This skill appears coherent and safe from a permissions perspective because it asks for no installs or credentials. Before using it: (1) avoid pasting sensitive PII or confidential numbers into prompts—when needed, redact or use placeholders; (2) if you want to avoid any chance of fabricated metrics, tell the agent to "先问清楚再写" (ask clarifying questions before drafting); (3) review any autogenerated numbers/claims and replace placeholders with real data; and (4) explicitly request the language/format desired (CN/EN/bilingual) to avoid unintended output. If you need stronger privacy guarantees, do not paste client names, financial amounts, or internal IDs into the editor.Like a lobster shell, security has layers — review code before you run it.
latestvk97934qjv96etqqfx40d3yw49984rz5f
License
MIT-0
Free to use, modify, and redistribute. No attribution required.