# AI use cases for Not AI (udaysharmadev/Not-Ai) Repo: [udaysharmadev/Not-Ai](https://github.com/udaysharmadev/Not-Ai) | Use cases: 4 | Page: https://dedupedai.com/use-cases/udaysharmadev-not-ai/ Not Ai is an Agent Skill and Python gate that helps rewrite and review prose for clarity, specificity, and voice while keeping facts, claims, and citations intact, and it does not disguise authorship. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Tighten a GitHub README draft before publishing: https://dedupedai.com/use-cases/udaysharmadev-not-ai/tighten-github-readme-draft/ - Revise a LinkedIn or social post for plain, specific wording: https://dedupedai.com/use-cases/udaysharmadev-not-ai/revise-linkedin-or-social-post/ - Check an academic draft and keep key statistics intact: https://dedupedai.com/use-cases/udaysharmadev-not-ai/check-academic-draft-protected-figures/ - Measure human-written rewrite pairs with the benchmark script: https://dedupedai.com/use-cases/udaysharmadev-not-ai/benchmark-human-rewrite-pairs/ ## How to tighten a GitHub README draft before publishing with Not AI Repo: [udaysharmadev/Not-Ai](https://github.com/udaysharmadev/Not-Ai) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Rewrite a project README for clarity and check it with the README gate profile. **Who it is for:** A developer about to publish a repo who has a rough README and wants it factual, plain, and free of marketing openers. **The problem:** Technical docs often read as vague or padded, and the developer needs them clear without changing what the project does. **The outcome:** A revised README draft with protected facts kept, plus advisory findings from the readme profile to review by hand. ### Steps 1. Run `npx skills add udaysharmadev/Not-Ai` to install the skill. 2. Save your README draft as readme-draft.txt in the repo folder. 3. Ask your agent to use the Not Ai skill on readme-draft.txt in readme mode, keeping all commands, file names, and numbers unchanged. 4. Clone the repo with git clone https://github.com/udaysharmadev/Not-Ai and move into the folder so the scripts are available. (inferred, check the README) 5. Run `python3 scripts/gate.py draft.txt --genre readme` on the revised text and review any findings. ### Prompt to paste into your agent ```text Using the udaysharmadev/Not-Ai skill, rewrite in readme mode, keep every command, file path, and number exactly as written, and return the revised file. Verify that the gate command `python3 scripts/gate.py --genre readme` completes without a hard failure. ``` ### Caveat The gate only blocks empty output and missing protected literals, so the editorial findings and the factual accuracy of the rewrite still need a human check. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/udaysharmadev-not-ai/tighten-github-readme-draft/ ## How to revise a LinkedIn or social post for plain, specific wording with Not AI Repo: [udaysharmadev/Not-Ai](https://github.com/udaysharmadev/Not-Ai) | Difficulty: Beginner | Time: 15 minutes | Basis: From the README Edit a short post for a social profile and check it against the linkedin or social profile. **Who it is for:** A professional posting on LinkedIn or X who has a draft that sounds generic and wants it to read like a person wrote it. **The problem:** Generic phrasing and stock openers weaken a post, but word swapping alone does not fix unclear purpose. **The outcome:** A tighter post that keeps the writer's claims and position, plus advisory findings on openers, rhythm, and vocabulary. ### Steps 1. Run `npx skills add udaysharmadev/Not-Ai` to install the skill. 2. Paste your draft into post.txt. 3. Ask your agent to rewrite post.txt with the Not Ai skill, keeping every claim and name as written. 4. Clone the repo and move into its folder. (inferred, check the README) 5. Run `python3 scripts/gate.py post.txt --genre linkedin` and review the advisory findings before posting. (inferred, check the README) ### Prompt to paste into your agent ```text Use the udaysharmadev/Not-Ai skill to revise for the linkedin profile, preserving every claim, name, and figure. Return the revised post. Verify that `python3 scripts/gate.py --genre linkedin` runs without a hard failure. ``` ### Caveat The README says the skill edits purpose, structure, and voice and does not disguise authorship, so the post should still reflect the writer's own position. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/udaysharmadev-not-ai/revise-linkedin-or-social-post/ ## How to check an academic draft and keep key statistics intact with Not AI Repo: [udaysharmadev/Not-Ai](https://github.com/udaysharmadev/Not-Ai) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Review a student or academic draft while protecting a required statistic from being altered. **Who it is for:** A student or researcher preparing a paper or report who needs the prose improved without changing a reported result. **The problem:** Edits can accidentally change a p-value or other exact literal, and the writer needs a check that catches it. **The outcome:** A revised draft plus a JSON gate report that fails if the protected literal is missing. ### Steps 1. Run `npx skills add udaysharmadev/Not-Ai` to install the skill. 2. Save the draft as draft.txt and note each exact literal that must stay, such as p = 0.03. 3. Ask your agent to revise draft.txt in academic style without changing any protected literal. 4. Clone the repo and move into its folder. (inferred, check the README) 5. Run `python3 plugins/not-ai/tools/gate.py draft.txt --genre academic --protect "p = 0.03" --json` and check the output. ### Prompt to paste into your agent ```text Using the udaysharmadev/Not-Ai skill, revise in academic style and keep the literal p = 0.03 unchanged. Return the revised text. Verify that `python3 plugins/not-ai/tools/gate.py --genre academic --protect "p = 0.03" --json` does not report a missing protected literal. ``` ### Caveat The gate does not claim factual fidelity or semantic equivalence, so the writer must confirm that every claim and citation still matches the sources. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/udaysharmadev-not-ai/check-academic-draft-protected-figures/ ## How to measure human-written rewrite pairs with the benchmark script with Not AI Repo: [udaysharmadev/Not-Ai](https://github.com/udaysharmadev/Not-Ai) | Difficulty: Advanced | Time: 60 minutes | Basis: From the README Compare original and rewritten text pairs using the repo's reproducible diagnostics. **Who it is for:** A researcher or editor who has human-provided original and rewritten pairs and wants repeatable diagnostics for a regression check. **The problem:** Judging whether a rewrite preserved content is hard to repeat by eye, and the team needs consistent numbers. **The outcome:** A benchmark report with token-overlap proxy, structural delta, readability delta, word-count change, and protected-literal preservation. ### Steps 1. Clone https://github.com/udaysharmadev/Not-Ai and move into the folder. (inferred, check the README) 2. Place your human-provided original and rewritten pairs in the fixture format the benchmark script expects, with metadata for the expected action. 3. Run `python3 scripts/benchmark.py` against the fixtures (inferred, check the script's help output for options). (inferred, check the README) 4. Read the token-overlap, structural, readability, and protected-literal results for each pair. 5. Treat the numbers as diagnostics and review any pair that changes meaning by hand. ### Prompt to paste into your agent ```text In the udaysharmadev/Not-Ai repo, run scripts/benchmark.py on the fixture pairs in and return the per-pair token-overlap, structural delta, readability delta, and protected-literal results. Verify that every pair reports a result and that no pair is labeled as a quality judgment. ``` ### Caveat The README says these figures are reproducible diagnostics, not proof that a rewrite is good or semantically identical. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/udaysharmadev-not-ai/benchmark-human-rewrite-pairs/