Pastewatch MCP — Secret Redaction

Prevents secrets from reaching your LLM provider. The agent works with placeholders, secrets stay local.

Source: https://github.com/ppiankov/pastewatch

Install

# macOS
brew install ppiankov/tap/pastewatch

# Linux (binary + checksum)
curl -fsSL https://github.com/ppiankov/pastewatch/releases/latest/download/pastewatch-cli-linux-amd64 \
  -o /usr/local/bin/pastewatch-cli
curl -fsSL https://github.com/ppiankov/pastewatch/releases/latest/download/pastewatch-cli-linux-amd64.sha256 \
  -o /tmp/pastewatch-cli.sha256
cd /usr/local/bin && sha256sum -c /tmp/pastewatch-cli.sha256
chmod +x /usr/local/bin/pastewatch-cli

Verify: pastewatch-cli version (expect 0.24.1+)

MCP Server Setup

mcporter config add pastewatch --command "pastewatch-cli mcp --audit-log /var/log/pastewatch-audit.log"
mcporter list pastewatch --schema  # 6 tools

Severity threshold: pastewatch-cli mcp --min-severity medium or set mcpMinSeverity in .pastewatch.json.

Agent Integration (one-command setup)

pastewatch-cli setup claude-code    # hooks + MCP + CLAUDE.md snippet + doctor check
pastewatch-cli setup cline          # MCP + hook instructions
pastewatch-cli setup cursor         # MCP + advisory

--severity aligns thresholds. --project for project-level config. Idempotent — safe to re-run.

MCP Tools

Tool	Purpose
`pastewatch_read_file`	Read file with secrets replaced by placeholders
`pastewatch_write_file`	Write file, resolving placeholders back locally
`pastewatch_check_output`	Verify text contains no raw secrets
`pastewatch_scan`	Scan text for sensitive data
`pastewatch_scan_file`	Scan a file
`pastewatch_scan_dir`	Scan directory recursively

Placeholder format: __PW_TYPE_N__ (changed from __PW{TYPE_N}__ in v0.19.7 for LLM proxy compatibility). Configurable via placeholderPrefix.

API Proxy — Last Line of Defense

Catches secrets from sub-agents that bypass hooks and MCP. Sits between your agent and the LLM provider, scanning all outbound requests.

v0.24.0+: Alert injection — when secrets are redacted, a [PASTEWATCH] alert is prepended to the API response so the agent gets immediate feedback. Disable with --no-alert.

pastewatch-cli proxy                    # scans all outbound API requests
pastewatch-cli proxy --port 9998 --upstream https://api.anthropic.com
pastewatch-cli proxy --port 9998 --upstream https://api.anthropic.com --no-alert  # silent redaction
pastewatch-cli proxy --forward-proxy http://corporate:8080  # chain with corporate proxy
pastewatch-cli proxy --severity medium --audit-log /var/log/pastewatch-proxy.log

Detected Secret Types (36)

Includes: AWS, Anthropic, OpenAI, Groq, Perplexity, Hugging Face, Workledger (wl_sk_), Oracul (vc_*_), npm, PyPI, RubyGems, GitLab, Telegram Bot, SendGrid, Shopify, DigitalOcean, Slack/Discord webhooks, Azure connections, GCP service accounts, SSH keys, JWTs, JDBC URLs, XML credentials, DB connections, emails, IPs, cards, and more.

Proxy + Chainwatch Chain (recommended for OpenClaw)

Stack both proxies for defense-in-depth:

OpenClaw → chainwatch:9999 (policy) → pastewatch:9998 (secrets) → api.anthropic.com

Setup as systemd services:

# 1. Pastewatch proxy (starts first)
cat > /etc/systemd/system/pastewatch-proxy.service << 'EOF'
[Unit]
Description=Pastewatch API Proxy (secret redaction)
After=network-online.target
Before=chainwatch-intercept.service
[Service]
Type=simple
ExecStart=/usr/local/bin/pastewatch-cli proxy \
  --port 9998 --upstream https://api.anthropic.com \
  --severity high --audit-log /var/log/pastewatch-proxy.log
Restart=always
RestartSec=3
MemoryMax=128M
[Install]
WantedBy=multi-user.target
EOF

# 2. Update chainwatch to forward to pastewatch (not Anthropic directly)
# Change --upstream from https://api.anthropic.com to http://localhost:9998

# 3. Enable and start
systemctl daemon-reload
systemctl enable pastewatch-proxy
systemctl start pastewatch-proxy
systemctl restart chainwatch-intercept

Rollback: Stop pastewatch-proxy, revert chainwatch --upstream to https://api.anthropic.com, restart chainwatch.

Guard — Block Secret-Leaking Commands

pastewatch-cli guard "cat .env"              # BLOCKED if .env has secrets
pastewatch-cli guard "psql -f migrate.sql"   # scans SQL file
pastewatch-cli guard "docker-compose up"     # scans env_files

Covers: shell builtins, DB CLIs (psql/mysql/mongosh/redis-cli/sqlite3), infra tools (ansible/terraform/docker/kubectl/helm), scripting (python/ruby/node/perl/php), file transfer (scp/rsync/ssh), pipe chains, subshells, redirects.

File Watcher

pastewatch-cli watch .              # continuous monitoring, real-time detection

Canary Tokens

pastewatch-cli canary generate --prefix myagent
pastewatch-cli canary verify
pastewatch-cli canary check --log /var/log/app.log

Encrypted Vault

pastewatch-cli --init-key                    # 256-bit key, mode 0600
pastewatch-cli fix --encrypt                 # → ChaCha20-Poly1305 vault
pastewatch-cli vault list | decrypt | export | rotate-key

Git History Scanning

pastewatch-cli scan --git-log
pastewatch-cli scan --git-log --range HEAD~50..HEAD --since 2025-01-01

Deduplicates by fingerprint. Output: text, json, sarif, markdown.

Org Posture Scanning

pastewatch-cli posture --org ppiankov           # scan all repos in org/user
pastewatch-cli posture --repos org/repo1,org/repo2
pastewatch-cli posture --compare previous.json  # trend tracking
pastewatch-cli posture --findings-only           # hide clean repos

Dashboard & Reports

pastewatch-cli dashboard                                  # aggregate across sessions
pastewatch-cli report --audit-log /var/log/pastewatch-audit.log
pastewatch-cli report --format json --since 2026-03-01T00:00:00Z

Configuration

pastewatch-cli init                         # generate .pastewatch.json
pastewatch-cli init --profile banking       # enterprise: JDBC, RFC 1918 IPs, service accounts
pastewatch-cli config                       # show resolved config
pastewatch-cli doctor                       # health check

Config cascade: /etc/pastewatch/config.json (admin) > project .pastewatch.json > ~/.config/pastewatch/config.json.

Features: sensitiveHosts, sensitiveIPPrefixes, xmlSensitiveTags, mcpMinSeverity, placeholderPrefix.

Detection Scope

30+ types: AWS keys/secrets, Anthropic/OpenAI/HuggingFace/Groq/Perplexity keys, DB connections, JDBC URLs, SSH keys, JWTs, emails, IPs, credit cards (Luhn), Slack/Discord webhooks, Azure, GCP service accounts, npm/PyPI/RubyGems/GitLab tokens, Telegram bot tokens, XML credentials, and more.

Gitignore-aware scanning (v0.21.0+). Deterministic regex. No ML. No API calls.

Limitations

Protects secrets from reaching LLM provider — does NOT protect prompt content or code structure
For full privacy, use a local model

This tool follows the Agent-Native CLI Convention. Validate with: clawhub install ancc && ancc validate .

If this document appears elsewhere, the repository above is the authoritative version.

pastewatch-mcp

Safety Notice

Copy this and send it to your AI assistant to learn