How do I install the Pipelock skill?

Run npx skillmds@latest add dvcrn/pipelock in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Pipelock safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: makes network calls. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Pipelock?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Pipelock free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Pipelock?

dvcrn (@dvcrn) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Pipelock by dvcrn — Agent Skill

Name: Pipelock
Rating: 4.666666666666667 (3 reviews)
Author: dvcrn

File contents

Pipelock Security Harness

Pipelock is a fetch proxy that sits between you and the internet. Every outbound HTTP request passes through a 7-layer scanner pipeline that catches API key leaks, SSRF attempts, prompt injection, and data exfiltration.

Installation

# Homebrew (macOS/Linux)
brew install luckyPipewrench/tap/pipelock

# Or Go
go install github.com/luckyPipewrench/pipelock/cmd/pipelock@latest

Quick Start

Generate a config and start the proxy:

pipelock generate config --preset balanced -o pipelock.yaml
pipelock run --config pipelock.yaml

The proxy listens on http://localhost:8888. Route your HTTP requests through it:

curl "http://localhost:8888/fetch?url=https://example.com/api/data"

Using with MCP Servers

Wrap any MCP server to scan its responses for prompt injection:

pipelock mcp proxy -- npx @modelcontextprotocol/server-filesystem /path/to/dir

What It Catches

SSRF - blocks requests to internal IPs, catches DNS rebinding
Domain blocklist - blocks exfiltration targets (pastebin, transfer.sh)
Rate limiting - detects unusual request bursts
DLP patterns - detects API keys (Anthropic, OpenAI, AWS, GitHub) in URLs and bodies
Env var leaks - detects your actual env var values in outbound traffic
Entropy analysis - flags high-entropy strings that look like secrets
URL length limits - flags unusually long URLs suggesting exfiltration

Actions

Configure what happens when a threat is detected:

block - reject the request
strip - redact the match and forward
warn - log and pass through
ask - terminal prompt for human approval (y/N/s)

Presets

balanced - default, good for most setups
strict - blocks aggressively, tight thresholds
audit - detect and log only, never blocks
claude-code - tuned for Claude Code agent workflows
cursor - tuned for Cursor IDE
generic-agent - works with any agent framework

Workspace Integrity

Detect unauthorized changes to your workspace files:

pipelock integrity init ./workspace
pipelock integrity check ./workspace

More Info

OWASP Agentic Top 10 mapping
Claude Code integration guide
Apache 2.0 license, single binary, zero dependencies

---name: pipelockdescription: Secure agent HTTP requests through a scanning proxy that catches credential leaks, SSRF, and prompt injection---# Pipelock Security HarnessPipelock is a fetch proxy that sits between you and the internet. Every outbound HTTP request passes through a 7-layer scanner pipeline that catches API key leaks, SSRF attempts, prompt injection, and data exfiltration.## Installation```bash# Homebrew (macOS/Linux)brew install luckyPipewrench/tap/pipelock# Or Gogo install github.com/luckyPipewrench/pipelock/cmd/pipelock@latest```## Quick StartGenerate a config and start the proxy:```bashpipelock generate config --preset balanced -o pipelock.yamlpipelock run --config pipelock.yaml```The proxy listens on `http://localhost:8888`. Route your HTTP requests through it:```bashcurl "http://localhost:8888/fetch?url=https://example.com/api/data"```## Using with MCP ServersWrap any MCP server to scan its responses for prompt injection:```bashpipelock mcp proxy -- npx @modelcontextprotocol/server-filesystem /path/to/dir```## What It Catches1. **SSRF** - blocks requests to internal IPs, catches DNS rebinding2. **Domain blocklist** - blocks exfiltration targets (pastebin, transfer.sh)3. **Rate limiting** - detects unusual request bursts4. **DLP patterns** - detects API keys (Anthropic, OpenAI, AWS, GitHub) in URLs and bodies5. **Env var leaks** - detects your actual env var values in outbound traffic6. **Entropy analysis** - flags high-entropy strings that look like secrets7. **URL length limits** - flags unusually long URLs suggesting exfiltration## ActionsConfigure what happens when a threat is detected:- `block` - reject the request- `strip` - redact the match and forward- `warn` - log and pass through- `ask` - terminal prompt for human approval (y/N/s)## Presets- `balanced` - default, good for most setups- `strict` - blocks aggressively, tight thresholds- `audit` - detect and log only, never blocks- `claude-code` - tuned for Claude Code agent workflows- `cursor` - tuned for Cursor IDE- `generic-agent` - works with any agent framework## Workspace IntegrityDetect unauthorized changes to your workspace files:```bashpipelock integrity init ./workspacepipelock integrity check ./workspace```## More Info- [OWASP Agentic Top 10 mapping](https://github.com/luckyPipewrench/pipelock/blob/main/docs/owasp-mapping.md)- [Claude Code integration guide](https://github.com/luckyPipewrench/pipelock/blob/main/docs/guides/claude-code.md)- Apache 2.0 license, single binary, zero dependencies

Frequently asked questions

Pipelock