agentleFS
Sign inStart free
JiamingShi-JsPhysJiamingShi-JsPhys/chelate-shell/skills/security-guard/SKILL.md

security-guard

Skill10 starsChanged 6 days agoMITView on GitHub

Screens text and Python sources before dispatch: prompt-injection markers in external text, and an AST walk for dangerous call sites (`eval`, `exec`, `os.system`, `subprocess.*`). A screen, not a sandbox.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    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:

    Run this exact form — do not read the source tree:

    line 25
  2. Distinctive:

    A screen, not a sandbox: Never describe a pass as "safe".

    line 54

≈ “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

security-guard

Flags known injection markers and dangerous call sites so a later stage can refuse the input. It reports risk; it cannot prove a file is safe.

  • Stage: infrastructure · Sandbox: no · Dependencies: stdlib only

Quick Execution

Run this exact form — do not read the source tree:

python run.py --root <WS> --action gate --input draft.md --mode auto
  • --mode {text,source,auto} picks the scanner; auto decides from the file.
  • --json emits the raw report.
  • The gate lands in <WS>/process/gates/security-guard_<ts>.json.

When to use

  • Untrusted text is about to enter the pipeline.
  • A Python file must be checked for dangerous calls before it runs.

When NOT to use

  • You need containment — that is the sandbox, not this skill.
  • The input is already trusted, first-party data.

Actions

--actionDoesExtra arguments
gateScreen and emit a gate result--input <path|text> [--mode text|source|auto] [--json]
scanScreen without gatingsame arguments

Hard constraints

  • A screen, not a sandbox. Never describe a pass as "safe".
  • Report the marker and its line, never a rewritten file.

Outputs

  • <WS>/process/gates/security-guard_<ts>.json — the gate result with every hit listed.

Failure handling

  • An unreadable input exits non-zero with a clear message.
  • A clean input is a pass with an explicit "no marker found" note.

Limitations

  • Marker and AST heuristics only; unknown attack shapes pass.
  • It never modifies the screened file.

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