agentleFS
Mohamed-Sherif-AliMohamed-Sherif-Ali/botter/.claude/skills/security-check/SKILL.md

security-check

Skill0 starsChanged 5 months agoMITView on GitHub

Audit a bot for the security mistakes that matter most before it goes live or gets handed to someone else: leaked or committed secrets, more permissions than it needs, risky handling of user input, and weak dependency hygiene. Use this when the user runs /security-check, or asks whether a bot is safe, secure, or ready to deploy or sell. Reports findings in plain language, fixes the safe ones, and walks the user through the rest.

  • Reads credentials

34 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use fix-bug, write-tests and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • JetBrainsSkill

    fix-bug

    End-to-end bug fix workflow for the Exposed project. Accepts a GitHub issue (#NUMBER), YouTrack issue (EXPOSED-NUMBER), or YouTrack URL (https://youtrack.jetbrains.com/issue/EXPOSED-NUMBER). Fetches the issue, creates a failing reproducer test, commits on a new branch, implements the fix, validates, and opens a PR. Use this skill whenever the user wants to fix a bug from an issue tracker, mentions a EXPOSED issue number, or references a GitHub issue to fix.

    Also in 3 projects that use this

  • SentrySkill

    write-tests

    Write tests following project conventions. Use when adding new tests or modifying existing ones.

    Also in 3 projects that use this

  • genomewalkerSkill

    learn

    Learn from web sources, generate SKILL.md, and store key patterns in soul memory. Trigger keywords: learn, skill-gen, web-learn, documentation, tutorial.

    Also in 3 projects that use this

  • GoogleSkill

    code-review

    Adversarial systems, concurrency, memory-model, and architecture code review protocol. Use this skill whenever asked to review code, PRs, diffs, or branches in Filament.

    Also in 6 projects that use this

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 26 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check a bot for the security problems that actually bite small bots, and tell the user plainly what to fix.

    line 8
  2. Distinctive:

    If no bot was named above, look in the bots/ folder, list the bots by name, and ask which one to check.

    line 23
  3. Distinctive:

    Run npm audit in the bot's folder and report anything serious in plain terms.

    line 54
  4. Distinctive:

    Keep it plain and beginner-friendly.

    line 67

“Common” is measured: kept by at least 6 in 10 of the 26 edited versions of this file we could read. “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

security-check

Check a bot for the security problems that actually bite small bots, and tell the user plainly what to fix. Read the bot-style base and the bot's platform style skill first so you know what a correct, minimal setup looks like.

Talk to the user as if they are not a programmer. Use plain language; explain why each finding matters, not just that it exists. Do not change behaviour while auditing; only fix the specific safe things below, and only after saying what you are doing.

The bot the user wants checked:

$ARGUMENTS

Work through these in order, and at the end give a short plain-language summary grouped by how urgent each finding is.

  1. If no bot was named above, look in the bots/ folder, list the bots by name, and ask which one to check. Wait. Note its platform and shape so the permission checks below apply correctly.
  2. Secrets, the most important check. A leaked secret is the worst thing that can happen to a bot.
    • Is the .env file listed in .gitignore? If not, that is how secrets get committed by accident.
    • Is .env actually tracked by git right now (check with git ls-files)? Was a secret ever committed in the past (scan the git history for the .env file or for token-shaped strings)? If either is true, treat the secret as leaked: it lives in git history even if deleted now, so it must be rotated, not just removed. Send the user to rotate-secret, and explain why deleting the file is not enough.
    • Is any real secret hardcoded in the source instead of read from .env? A token or webhook URL pasted into the code is the same as committing it.
    • Does .env.example contain only placeholders, no real values?
    • Does the code print or log the secret anywhere (for example, logging the token or the whole environment)? Secrets must never reach the logs, especially on a shared host.
  3. Least privilege: the bot should ask for no more power than its job needs.
    • Discord full bot: are the requested intents the minimum for the job, with any privileged one actually used? Is the invite link's permission set minimal? Flag anything broad it does not use.
    • Telegram: is privacy mode left on unless the bot genuinely needs to read all group messages? Turning it off is a real expansion of what the bot can see, so it should be deliberate.
    • A poster (either platform) should hold only its send-only secret and nothing more.
  4. Handling of incoming input (full bots only). If the bot acts on what users send, make sure it never feeds that text into something dangerous: running a shell command, evaluating code, reading or writing a file path built from the message, or building a query from it. Beginner bots usually do none of this; if this one does, flag it clearly, because that is how a bot gets taken over.
  5. Dependency hygiene.
    • Are dependencies pinned to specific versions rather than loose wildcards, so a deploy cannot silently pull different code?
    • Run npm audit in the bot's folder and report anything serious in plain terms.
    • Does the bot carry only the dependencies its shape needs (for example, a poster should not depend on discord.js or grammY)?
  6. Data file hygiene, if the bot saves data. The data file should be git-ignored and left out of any package, so one person's saved data never ships to someone else.
  7. Fix and report. Offer to do the safe fixes yourself: add .env or a data file to .gitignore, pin a loose dependency, remove a line that logs a secret. For anything that needs the user or a judgement call, rotating a leaked secret, narrowing intents, turning privacy mode back on, walk them through it. End with a short summary: what is fine, what you fixed, and what still needs them, ordered most urgent first.

Keep it plain and beginner-friendly. When unsure whether something is a real risk, say so honestly rather than raising false alarms or glossing over it.

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.