agentleFS
Sign inStart free
Williamunpalatable7706Williamunpalatable7706/ai-life-skills-toolkit/skills/core/use-private-profile-safely/SKILL.md

use-private-profile-safely

Skill0 starsChanged 2 months agoMITView on GitHub

Handle optional local user preferences and encrypted personal-profile fields without exposing them to Git, logs, screenshots, or unrelated tasks. Use when a user asks to create, store, retrieve, or apply personal information through this toolkit.

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 8 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the minimum personal information needed for the current task.

    line 8
  2. Distinctive:

    Use known non-sensitive context and safe, reversible defaults before asking for input.

    line 13
  3. Very specific:

    Put non-sensitive preferences such as operating system, preferred browser, timezone, teaching style, and desired access supports in a local …

    line 22
  4. Very specific:

    Put genuinely personal details in the encrypted vault created by personal-vault/vault.py.

    line 25
  5. Distinctive:

    Keep passwords, passkeys, MFA/recovery codes, authentication tokens, full government-ID numbers, and complete payment or bank numbers in a …

    line 27
  6. Distinctive:

    Never ask the user to paste the master passphrase into chat or a command-line argument.

    line 37
  7. Distinctive:

    Never inspect the entire vault merely to answer a question about one field.

    line 39
  8. Very specific:

    Prefer list for field names and get without --reveal when existence is enough.

    line 40
  9. Distinctive:

    Do not repeat personal values in logs, screenshots, generated examples, issue reports, commits, or final summaries.

    line 42
  10. Distinctive:

    Treat stored dates and statuses as potentially stale.

    line 44
  11. Distinctive:

    Run from the repository root:

    line 49

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Use a Private Profile Safely

Use the minimum personal information needed for the current task. A user profile is not blanket permission to inspect every stored field.

Keep momentum

Use known non-sensitive context and safe, reversible defaults before asking for input. Bundle non-blocking questions and continue useful work. Interrupt only when a missing choice would change the result, privacy, cost, authority, safety, or something hard to undo — and never treat convenience as permission to reveal private data. Never repeat an answered question or request access to fields the task does not need.

Choose the storage tier

  1. Put non-sensitive preferences such as operating system, preferred browser, timezone, teaching style, and desired access supports in a local copy of config/preferences.example.json named config/preferences.local.json.
  2. Put genuinely personal details in the encrypted vault created by personal-vault/vault.py.
  3. Keep passwords, passkeys, MFA/recovery codes, authentication tokens, full government-ID numbers, and complete payment or bank numbers in a trusted password manager or operating-system credential store instead.

Both the local preferences file and encrypted vault patterns are ignored by Git. Never weaken or remove those ignore rules while real personal data exists.

Access boundary

  • Obtain explicit permission for the exact profile field needed.
  • Never ask the user to paste the master passphrase into chat or a command-line argument. The user enters it directly into the local hidden prompt.
  • Never inspect the entire vault merely to answer a question about one field.
  • Prefer list for field names and get without --reveal when existence is enough. Reveal a value only on a private screen and only when required.
  • Do not repeat personal values in logs, screenshots, generated examples, issue reports, commits, or final summaries.
  • Treat stored dates and statuses as potentially stale. Recheck live state before an account, medical, legal, financial, renewal, or submission action.

Local commands

Run from the repository root:

python personal-vault/vault.py list private/profile.vault.json
python personal-vault/vault.py get private/profile.vault.json field.name

The user must run any command that prompts for the master passphrase. If the task cannot continue without exposing a protected value, stop and explain the safe handoff instead of bypassing encryption.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free