Skip to content
TakoAPI
Back to skills

messageguard

by andrewandrewsen

CLI Utilities
248Downloads reported by clawskills.sh
2Views
0 likesSign in to like this skill
—Not yet rated
Your rating

Brief

**Purpose**: MessageGuard filters outgoing text to prevent secret leaks and sensitive data exposure.

Description

What This Skill Does

MessageGuard scans outgoing messages for sensitive data patterns before they are sent. It supports three response modes: masking matched content, blocking the message entirely, or generating a warning. Built-in patterns cover common secrets like API keys, AWS credentials, and JWTs, with support for custom regex patterns.

Catches secret leaks at the point of sending rather than after the fact, with no third-party dependencies required.

When to Use It

  • Blocking API keys from being pasted into shared chat logs
  • Masking JWT tokens before sending debug output to teammates
  • Warning when AWS credentials appear in outgoing messages
  • Logging credential detections for compliance audits
  • Enforcing secret hygiene in automated messaging pipelines

Example Workflow

INPUT

User asks: send this debug log to the team channel

AGENT
  1. 1

    Intercepts the outgoing message before delivery

  2. 2

    Runs the content against configured regex patterns

  3. 3

    Detects an AWS access key pattern match

  4. 4

    Applies the configured mode action (mask, block, or warn)

  5. 5

    Writes a structured JSON detection entry to the configured log path

OUTPUT

Message is blocked or masked and the detection is recorded in the log file

Installation

clawhub install andrewandrewsen/messageguard

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links