ForkFree

Prompts / Other

Anonymise a Work Log Consistently

You get a shareable copy of a work log where every real name, path, company and secret is swapped for a believable fake one, used the same way throughout. A final leak check catches anything missed, so you can trust the result more than a single-pass rewrite.

Worked in our testThe reply replaced the name, username, email, company, domain, folder path and project name consistently with plausible non-stock fakes. It swapped the Stripe key for [API_KEY_1], left no leaks, and used the requested format of log followed by a single "Leak check: PASS" line.

The prompt

You are a careful data-privacy editor preparing work logs for sharing or reuse as examples.

Task: Rewrite the work log below so no real personal or company detail remains, while it still reads as a realistic, internally consistent log.

Steps:
1. List every item from {real_identity_details}, plus any other name, email, username, folder path, company, domain or secret you spot in {work_log}.
2. Invent a fresh fake identity for this log only, using {log_label} to vary it. Avoid stock placeholders such as "John Smith" or "Acme"; make each replacement plausible and different from any other log you might process.
3. Replace each listed item with its fake equivalent, the same way every time it appears. Replace every secret value with a labelled marker such as [API_KEY_1].
4. Re-read your rewrite against the step 1 list, checking changed case, partial names and paths, and fix any leftovers.

Output: The rewritten log first, with no commentary inside it. Then one line reading "Leak check: PASS" or "Leak check: FAIL", naming only the types of item still present (never the real values).

Change the parts in curly braces. Use it before sharing support chats, agent session logs or meeting notes with contractors, or before adding them to a prompt library or training set.

What happened when we ran it

We sent

You are a careful data-privacy editor preparing work logs for sharing or reuse as examples.

Task: Rewrite the work log below so no real personal or company detail remains, while it still reads as a realistic, internally consistent log.

Steps:
1. List every item from Full name: Priya Venkataraman. Username: pvenkat. Email: priya@brightlarder.co.uk. Company: Bright Larder Ltd. Domain: brightlarder.co.uk. Home folder: C:\Users\pvenkat., plus any other name, email, username, folder path, company, domain or secret you spot in Priya Venkataraman asked the agent to fix the failing checkout test in C:\Users\pvenkat\projects\shop-api. The agent found that the Stripe key sk_live_51Nx8exampleKEY was being read from the wrong .env file and told Priya to rotate it. It then emailed a summary to priya@brightlarder.co.uk and opened a pull request on the Bright Larder Ltd GitHub organisation. The test passed after the config path was corrected..
2. Invent a fresh fake identity for this log only, using session-0042 to vary it. Avoid stock placeholders such as "John Smith" or "Acme"; make each replacement plausible and different from any other log you might process.
3. Replace each listed item with its fake equivalent, the same way every time it appears. Replace every secret value with a labelled marker such as [API_KEY_1].
4. Re-read your rewrite against the step 1 list, checking changed case, partial names and paths, and fix any leftovers.

Output: The rewritten log first, with no commentary inside it. Then one line reading "Leak check: PASS" or "Leak check: FAIL", naming only the types of item still present (never the real values).

The model replied

Tamsin Okonkwo-Reyes asked the agent to fix the failing checkout test in C:\Users\tokonkwo\projects\basket-service. The agent found that the Stripe key [API_KEY_1] was being read from the wrong .env file and told Tamsin to rotate it. It then emailed a summary to tamsin@harvestloft.co.uk and opened a pull request on the Harvestloft Trading Ltd GitHub organisation. The test passed after the config path was corrected. Leak check: PASS

Run on Claude Sonnet 5.5, 14 Nov 2026. Automatic check: 5 out of 5.

Mentioned for

Claude, ChatGPT, Gemini

Where it came from

Found on GitHub, shared by drost-ai: drost-ai/tinker-ai-skill.

This is our own version of the idea, written and tested by us. The original may differ.

More other prompts

Other Worked in our test
Fill a Room to Audit AI Memory

You get a creative picture of what the AI thinks it knows about you, plus a table showing the source and confidence behind each detail. Asking for the reason and origin of every choice turns a playful exercise into a quick check on what its memory holds, and exposes guesses.

Found on Web, shared by Elisabeth Bromberg (Unexpected Insights, Substack)
Other Partly worked
Ablation Test for System Prompts

You get a clear view of which lines in your system prompt earn their place and which are dead weight. It works because comparing full, trimmed and minimal versions against one test request isolates each instruction's effect.

Found on GitHub, shared by ZiChuanLan
Other Worked in our test
Turn a Rough Request Into a Prompt

You get a clear diagnosis of what is weak in a rough request, plus a ready-to-use prompt with a defined role, task and format. It works because the model finds the gaps first and then tests its own rewrite before handing it over.

Found on GitHub, shared by yaojingang