De-AI-ify
v1.0.0Remove AI-generated jargon and restore human voice to text
⭐ 6· 3.5k·25 current·26 all-time
by@itsflow
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description (remove AI jargon and restore human voice) match the runtime instructions: the SKILL.md describes reading a source file, producing a '-HUMAN' copy, applying transformations, and producing a change log. Nothing requested (no env vars, no binaries) appears unnecessary for a file-editing text transformation.
Instruction Scope
The instructions explicitly tell the agent to 'Read original file' and create a copy with a '-HUMAN' suffix. That is expected for a file-editing skill, but it means the agent will read whatever file path the user supplies — including potentially sensitive system files if the user provides them. The instructions do not transmit data externally or request additional context beyond the file, and they do not instruct reading other environment variables or unrelated paths.
Install Mechanism
No install spec and no code files (instruction-only). This is the lowest-risk install model: nothing is downloaded or written by an installer.
Credentials
No environment variables, credentials, or config paths are requested. The lack of secret/credential access is proportionate to the described text-editing function.
Persistence & Privilege
The skill is not flagged as always: true and does not request persistent system changes. It will operate only when invoked. Note: the platform default allows autonomous invocation, but there are no other indicators (no persistent installs or credential writes) that increase privilege.
Assessment
This skill is internally consistent for editing files, but be cautious when using it: 1) Only pass file paths you intend to edit — do not point it at system files, private keys, or other secrets. 2) Because it writes a new file with a '-HUMAN' suffix, review the output and change log before publishing or replacing originals. 3) The skill author and source are unknown and there is no homepage; if provenance matters for your workflow, prefer skills from known authors or test on non-sensitive sample files first. 4) The skill does not transmit data externally or request credentials, so the main risk is accidental exposure from supplying sensitive file paths.Like a lobster shell, security has layers — review code before you run it.
latestvk977dd524rfr6ry06qkm87085h7zv6y7
License
MIT-0
Free to use, modify, and redistribute. No attribution required.