CommunityCoding & Developmentgithub.com

Tchapwo1/tcha-claude-skills

In-line credential and secret redaction guard for Claude. Intercepts sensitive strings (AWS access keys, OpenAI/Anthropic API tokens, database URIs, private keys, JWTs) before they are sent to LLM APIs, logged to terminals, or committed to git.

What is tcha-claude-skills?

tcha-claude-skills is a Claude Code agent skill that in-line credential and secret redaction guard for Claude. Intercepts sensitive strings (AWS access keys, OpenAI/Anthropic API tokens, database URIs, private keys, JWTs) before they are sent to LLM APIs, logged to terminals, or committed to git.

Works with✓Claude Code✓Codex CLI~Cursor✓Antigravity
npx skills add https://github.com/Tchapwo1/tcha-claude-skills/tree/HEAD/.claude/skills/tcha-secret-shield

Installed? Explore more Coding & Development skills: steipete/bluebubbles, steipete/eightctl, steipete/blucli · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

Tcha Secret Shield

A deterministic, zero-latency credential sanitizer and redaction guard for Claude Code, Claude Desktop, and agentic workflows.

Capabilities & Scopes

  • Zero-Exposure Interception: Scans files, prompt inputs, and tool outputs before transmission.
  • Pattern Matching & Entropy Analysis:
    • Cloud Providers: AWS (AKIA...), GCP Service Accounts, Azure Client Secrets.
    • LLM APIs: Anthropic (sk-ant-...), OpenAI (sk-...), Cohere, Mistral.
    • Infrastructure: Stripe secret keys (sk_live_...), GitHub PATs (ghp_...), npm tokens (npm_...).
    • Cryptographic Assets: PEM/RSA private keys, SSH keys (id_rsa, id_ed25519).
    • Database URIs: PostgreSQL, MySQL, Redis, MongoDB connection strings with embedded passwords.
  • Deterministic Token Replacement: Replaces sensitive data with cryptographic placeholders (e.g. [REDACTED_TCHA_AWS_SECRET_<HASH>]) to preserve structural integrity without leaking secrets.

When to Activate

Trigger this skill whenever:

  1. Claude reads or displays local configuration files (.env, credentials.json, config.yaml, id_rsa).
  2. An agent prepares a git commit or pushes code to an external repository.
  3. Untrusted outputs are passed between multi-agent subtasks.

Command Line Integration

Execute directly in Claude Code terminal:

npx tcha-protection sanitize --file .env
# Or pipe input through the sanitizer
cat config.json | npx tcha-protection sanitize

Security Guarantees

  • Operates 100% locally on your machine.
  • Zero external network requests during redaction.
  • Latency overhead: <0.2ms.

Individual skills in this repo

This repo contains 2 individual skills — each has its own dedicated page.

Related Skills