Dev Security Audit

What it does

When your agent uses it

• The user suspects their machine may be compromised
• Wants to check for exposed secrets
• Asks about supply chain attacks
• Wants a full security audit of their development environment

Example prompts

• “am I compromised”
• “check my security”
• “scan for leaked keys”
• “/dev-security-audit”

Requirements

• Docker

Workflow steps

5 steps, taken from the step headings in SKILL.md.

1. 1Supply Chain IoC Dispatch
2. 2Credential & Secret Exposure Scan
3. 3Application & Browser Data Scan
4. 4Persistence & Backdoor Check
5. 5Report Generation

What it can do on your machine

Read from SKILL.md and the folder at commit c9a2036. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • aws
  • docker
  • osascript

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use aws and docker, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Dev Security Audit loads about 3.6k tokens when it runs, and up to ~14k if it reads all its reference files. Until then it costs about 144 tokens; SKILL.md has 1,195 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~144

When it runs · the whole SKILL.md, loaded when a task matches

~3.6k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~14k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:85

  - AWS (`~/.aws/credentials`, `~/.aws/config`)

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:88

  - Kubernetes (`~/.kube/config`, `~/.kube/custom-contexts/`)

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:90

  - Docker (`~/.docker/config.json`)

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:93

  - SSH keys (`~/.ssh/`)

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:97

  - npm (`~/.npmrc`)

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:98

  - GPG keys (`~/.gnupg/private-keys-v1.d/`)

• NoteMentions a .env fileSKILL.md:102

  - `.env` files (`find ~ -maxdepth 5 \( -name ".env" -o -name ".env.*" \) 2>/dev/null | grep -v node_modules | grep -v .g

• NoteMentions a .env fileSKILL.md:117

  to extract tokens from shell history and .env files:

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:176

  Browsers store Login Data, Cookies, and Local Storage accessible to user-space processes:

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:187

  Note: Chrome's Login Data is encrypted via macOS Keychain. Under RCE, the attacker could potentially decrypt it during a

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from sd0xdev/sd0x-harness at commit c9a2036, republished under its MIT licence (© sd0xdev). 1,195 words, ~3,553 tokens.

name

dev-security-audit

description

Comprehensive developer workstation security audit — scans for exposed credentials, compromised application data, persistence mechanisms, and supply chain attack indicators. Use this skill whenever the user suspects their machine may be compromised, wants to check for exposed secrets, asks about supply chain attacks, or wants a full security audit of their development environment. Also triggers on: 'am I compromised', 'check my security', 'scan for leaked keys', 'credential audit', 'supply chain attack', 'supply chain check', 'check if I was hacked'.

sequenceDiagram
    participant U as User
    participant C as Claude
    participant S as System
    C->>S: Phase 0: Supply Chain IoC Dispatch (case-driven)
    C->>S: Phase 1: Credential & Secret Exposure Scan
    C->>S: Phase 2: Application & Browser Data Scan
    C->>S: Phase 3: Persistence & Backdoor Check
    C->>U: Phase 4: Report Generation & Remediation Plan

OpenAI:           sk-[a-zA-Z0-9_-]{20,}
Anthropic:        sk-ant-[a-zA-Z0-9_-]{20,}
GitHub Classic:   gh[posur]_[a-zA-Z0-9]{20,}
GitHub Fine-grain: github_pat_[a-zA-Z0-9_]{20,}
GitLab PAT:       glpat-[a-zA-Z0-9_-]{20,}
AWS Access:       AKIA[A-Z0-9]{16}
AWS Temp:         ASIA[A-Z0-9]{16}
HuggingFace:      hf_[a-zA-Z0-9]{20,}
npm:              npm_[a-zA-Z0-9]{20,}
Docker Hub:       dckr_pat_[a-zA-Z0-9_-]{20,}
Slack:            xox[bsrp]-[a-zA-Z0-9-]{20,}
Stripe:           [rs]k_live_[a-zA-Z0-9]{20,}
Firebase:         AIza[a-zA-Z0-9_-]{30,}
JWT:              eyJ[a-zA-Z0-9_-]+\.eyJ[a-zA-Z0-9_-]+\.[a-zA-Z0-9_-]+
Vercel:           vercel_[a-zA-Z0-9_-]{20,}
Supabase:         sbp_[a-zA-Z0-9]{20,}

find ~/Library/Application\ Support/*/Local\ Storage/leveldb -maxdepth 0 2>/dev/null

# Chrome profiles
ls ~/Library/Application\ Support/Google/Chrome/*/Login\ Data 2>/dev/null
# Firefox
ls ~/Library/Application\ Support/Firefox/Profiles/*/logins.json 2>/dev/null
# Arc
ls ~/Library/Application\ Support/Arc/*/Login\ Data 2>/dev/null

ls ~/Library/Keychains/ 2>/dev/null

# User LaunchAgents (most common persistence vector)
ls ~/Library/LaunchAgents/

# System LaunchDaemons (requires root, less likely for user-space attack)
ls /Library/LaunchDaemons/

# Non-Apple launchctl services
launchctl list | grep -v com.apple

# Cron jobs
crontab -l

# Login items
osascript -e 'tell application "System Events" to get the name of every login item' 2>/dev/null

crontab -l
ls /etc/cron.d/
systemctl list-unit-files --type=service | grep -v disabled

schtasks /query /fo LIST /v
sc query type=service state=all | findstr /v /i "Microsoft Windows"
reg query "HKCU\Software\Microsoft\Windows\CurrentVersion\Run"

ls -la /usr/local/bin/ | head -30
# List executables modified within a suspected attack window
# Use dates from Phase 0 case match (attack_window.start/end) or user-provided window
WINDOW_START="${ATTACK_WINDOW_START:?Set ATTACK_WINDOW_START from Phase 0 case match}"
WINDOW_END="${ATTACK_WINDOW_END:?Set ATTACK_WINDOW_END from Phase 0 case match}"
find /usr/local/bin -type f -newermt "$WINDOW_START" ! -newermt "$WINDOW_END" 2>/dev/null | while read f; do
  file "$f" 2>/dev/null | grep -q "executable" && echo "SUSPECT: $f ($(stat -f '%Sm' "$f" 2>/dev/null || stat -c '%y' "$f" 2>/dev/null))"
done

umask 077
REPORT_PATH="${TMPDIR:-/tmp}/security-audit-$(hostname -s)-$(date +%Y%m%d-%H%M%S).md"
touch "$REPORT_PATH" && chmod 600 "$REPORT_PATH"
# Write report content to $REPORT_PATH
echo "Report saved to: $REPORT_PATH"

# Security Audit Report

## Summary
- Scan date: YYYY-MM-DD
- Platform: macOS/Linux/Windows
- Supply Chain IoC: [per-case status from Phase 0, or "No active cases matched"]
- Total findings: N (N critical, N high, N medium, N low)

## Supply Chain Status
[Per-case results table: case_id | status | confidence]

## Critical Findings (Immediate Action Required)
| # | Category | Item | Path | Action |

## High Findings (Action Within 24h)
| # | Category | Item | Path | Action |

## Medium Findings (Evaluate & Monitor)
| # | Category | Item | Path | Action |

## Recommended Action Plan
### Tier 0 — Immediately (minutes)
### Tier 1 — Today (hours)
### Tier 2 — This Week
### Tier 3 — Contingency Triggers

## What Was NOT Found (Good News)
[List of categories that came back clean]

© sd0xdev, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 5 other files (references) in skills/dev-security-audit of sd0xdev/sd0x-harness.

• SKILL.md
• references/cases/README.md
• references/cases/apifox-2026-03.md
• references/cases/axios-2026-03.md
• references/remediation.md
• references/scan-targets.md

Open the folder on GitHubat commit c9a2036