cso

Audits codebases for secrets, supply chain, CI/CD, LLM, and OWASP vulnerabilities.

107|7|Updated Mar 29, 2026
One-click install
npx skills add https://github.com/no-session/pstack --skill cso-no-session
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cso
Source: https://github.com/no-session/pstack/tree/main/cso
Command: npx skills add https://github.com/no-session/pstack --skill cso-no-session

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Solo founders and small teams rarely run real security reviews, leaving leaked secrets in git history, vulnerable dependencies, misconfigured CI/CD pipelines, and unverified webhooks undetected until a breach happens. ## Core Features & Use Cases - Infrastructure-first audit: Scans git history for leaked credentials, checks dependency supply chains, CI/CD pipeline security, Docker/IaC misconfigurations, and webhook signature verification. - LLM and skill supply chain security: Detects prompt injection vectors, unsanitized LLM output, unvalidated tool calls, and malicious patterns in installed Claude Code skills. - Two audit modes with confidence gating: Daily mode reports only 8/10-confidence findings for zero noise; comprehensive mode lowers the bar to 2/10 for monthly deep scans, with trend tracking across runs. - Use Case: Before shipping a new SaaS feature, run a scoped audit on the current branch diff to catch hardcoded API keys, unpinned GitHub Actions, and missing webhook signature checks, then receive a Security Posture Report with severity ratings and remediation plans. ## Quick Start Ask the AI to run a full security audit of this repository with /cso and produce a severity-rated Security Posture Report.

Frequently Asked Questions about cso

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a security audit on my codebase with Claude Code?

Invoke /cso with no flags for a full daily audit across all 15 phases with an 8/10 confidence gate. Use /cso --comprehensive for a monthly deep scan, or scope flags like --infra, --code, --owasp, or --diff to narrow the audit.

What does the cso skill check beyond npm audit?

It scans git history for leaked credentials, checks install scripts in production dependencies, verifies lockfile integrity, audits CI/CD workflows for unpinned actions and script injection, and inspects webhooks, Dockerfiles, IaC, and LLM-specific vulnerabilities.

Can I audit only the changes on my current branch?

Yes, the --diff flag constrains every phase to files and configs changed on the current branch versus the base branch. It combines with any scope flag and with --comprehensive mode.

Does the security audit support Python, Go, and Ruby projects?

Yes, Phase 0 detects Node/TypeScript, Ruby, Python, Go, Rust, JVM, PHP, and .NET stacks plus frameworks like Next.js, Django, Rails, and Gin. Detection sets scan priority, and a catch-all pass still covers undetected languages.

Why does the daily audit report fewer findings than comprehensive mode?

Daily mode applies an 8/10 confidence gate so only high-confidence, verified findings are reported, eliminating noise. Comprehensive mode lowers the bar to 2/10, surfacing more speculative findings for monthly deep review.

Does the audit make code changes or live requests to my endpoints?

No, the skill never modifies code and never makes live HTTP requests. It produces a Security Posture Report with findings, severity ratings, and remediation plans, verifying webhooks by tracing code only.