De Ai Ify
PassAudited by ClawScan on May 10, 2026.
Overview
This instruction-only writing skill appears purpose-aligned and shows no code, credentials, network calls, or persistence, but users should review file edits and not over-rely on its offline, accuracy, or fact-preservation claims.
This looks safe to use as a scoped document-rewriting prompt. Only run it on files you intend the agent to read, confirm your platform's data handling if the text is confidential, verify any added facts, and check the package version/source because the metadata is slightly inconsistent.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The selected document's contents will be processed and a new file may be created next to it.
The skill directs the agent to read a user-specified document and write a revised file. This is expected for an editing skill, but users should choose paths carefully.
"/de-ai-ify <file_path>" ... "Read original file" ... "Create \"-HUMAN.md\" file"
Run it only on documents you intend to process, and review the generated file before using or publishing it.
It may be harder to confirm exactly which package version or source was reviewed.
The registry metadata lists a different slug and the SKILL front matter lists version 2.0.0, while _meta.json lists version 1.0.0. With no code or install spec, this is mainly a provenance/version clarity issue.
"slug": "de-ai-ify", "latest": { "version": "1.0.0"Verify the package source and version before relying on the skill, especially if installing it outside the registry flow.
Users may process sensitive or important text believing there is stronger offline or accuracy assurance than the artifacts prove.
The package is instruction-only and provides no reviewed implementation or test data to substantiate deterministic offline processing or accuracy claims. The wording may overstate privacy or reliability.
"No API calls needed - Pure pattern matching, works offline" ... "92% accuracy"
Treat it as a prompt-based editing aid unless you have confirmed your OpenClaw runtime's data handling and the skill author's evidence.
The rewritten document could accidentally change or add factual claims even when the user expected only a style edit.
The skill promises fact preservation but also describes adding specific examples, which can introduce unsupported names, numbers, or claims.
"Keeps your facts, structure, and key points" ... "Added 5 specific examples (replaced vague references)"
Manually fact-check any added examples, statistics, company names, or concrete claims before publishing.