# Perseus Specialist

> Run all specialist deep-dive skills in parallel for comprehensive analysis

- **Type:** Skill
- **Install:** `agentstack add skill-kaivyy-perseus-all`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [kaivyy](https://agentstack.voostack.com/s/kaivyy)
- **Installs:** 0
- **Category:** [Security](https://agentstack.voostack.com/c/security)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [kaivyy](https://github.com/kaivyy)
- **Source:** https://github.com/kaivyy/perseus/tree/main/skills/perseus/specialists/all

## Install

```sh
agentstack add skill-kaivyy-perseus-all
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Perseus Specialist (All Deep-Dive Skills)

## Context & Authorization

**IMPORTANT:** This skill runs all specialist security analyses on the **user's own codebase**. This is defensive security testing to find vulnerabilities before attackers do.

**Authorization:** The user owns this codebase and has explicitly requested this comprehensive analysis.

---

## Overview

This skill orchestrates ALL 8 specialist skills in parallel for maximum coverage and efficiency.

**When to Use:**
- After `/scan` to get deep analysis on all detected components
- When you want comprehensive specialist coverage without running `/start`
- As a standalone deep-dive after initial assessment

**Goal:** Run all specialists simultaneously for thorough security analysis.

## Engagement Mode Compatibility

| Mode | Orchestrator Behavior |
|------|------------------------|
| `PRODUCTION_SAFE` | Run specialists with passive-first and low-rate verification constraints |
| `STAGING_ACTIVE` | Allow active specialist checks with throttling |
| `LAB_FULL` | Allow broad dynamic specialist verification in isolated lab |
| `LAB_RED_TEAM` | Allow chain-based specialist simulation with strict kill-switches |

## Safety Gates (Required)

1. Read `deliverables/engagement_profile.md` before launching specialists.
2. Default to `PRODUCTION_SAFE` if engagement mode is not available.
3. Propagate mode and rate limits to each specialist task.
4. Stop all specialists if any run reports `ABORTED-SAFETY`.

## Specialists Included

| Skill | Coverage | Output |
|-------|----------|--------|
| `perseus-api` | OWASP API Top 10, GraphQL, WebSocket | `api_security_analysis.md` |
| `perseus-injection` | NoSQL, LDAP, XPath, SSTI, Command | `injection_deep_analysis.md` |
| `perseus-crypto` | JWT, Hashing, Encryption, Secrets | `crypto_security_analysis.md` |
| `perseus-supply-chain` | CVEs, Dependencies, Licenses | `supply_chain_analysis.md` |
| `perseus-file` | Path Traversal, Upload, XXE | `file_security_analysis.md` |
| `perseus-logic` | Race Conditions, Business Logic | `business_logic_analysis.md` |
| `perseus-client` | DOM XSS, Prototype Pollution | `client_side_analysis.md` |
| `perseus-config` | Headers, CORS, Cookies, TLS | `config_security_analysis.md` |

## Execution Instructions

### Step 0: Mode & Scope Alignment

- Load mode/scope/limits from `deliverables/engagement_profile.md`.
- Respect `deliverables/verification_scope.md` when present.
- Announce mode before launching specialists.

### Step 1: Announce Start

```
"Running all Perseus specialist skills in parallel..."
"This provides deep-dive analysis across 8 security domains."
```

### Step 2: Launch All Specialists in Parallel

Use a single message with 8 parallel `Task` tool calls:

```
Parallel Tasks:
1. Task: "Run API security specialist" -> Skill: perseus-api
2. Task: "Run injection specialist" -> Skill: perseus-injection
3. Task: "Run crypto specialist" -> Skill: perseus-crypto
4. Task: "Run supply chain specialist" -> Skill: perseus-supply-chain
5. Task: "Run file security specialist" -> Skill: perseus-file
6. Task: "Run business logic specialist" -> Skill: perseus-logic
7. Task: "Run client-side specialist" -> Skill: perseus-client
8. Task: "Run config specialist" -> Skill: perseus-config
```

### Step 3: Wait for Completion

Wait for all 8 specialists to complete their analysis.

### Step 4: Summarize Results

```
"Specialist analysis complete!"

Summary:
- API Security: X findings
- Injection: X findings
- Cryptography: X findings
- Supply Chain: X findings
- File Security: X findings
- Business Logic: X findings
- Client-Side: X findings
- Configuration: X findings

Total: X findings across 8 domains

"All reports saved to deliverables/"
```

## Output Structure

After completion, `deliverables/` will contain:

```
deliverables/
├── api_security_analysis.md
├── injection_deep_analysis.md
├── crypto_security_analysis.md
├── supply_chain_analysis.md
├── file_security_analysis.md
├── business_logic_analysis.md
├── client_side_analysis.md
└── config_security_analysis.md
```

## When to Use Each Specialist Individually

| If You Need | Run |
|-------------|-----|
| Only API analysis | `/api` |
| Only injection deep-dive | `/injection` |
| Only crypto audit | `/crypto` |
| Only dependency check | `/supply-chain` |
| Only file/upload security | `/file` |
| Only business logic | `/logic` |
| Only client-side | `/client` |
| Only config hardening | `/config` |
| **All of the above** | `/specialist` |

## Integration with Core Skills

```
Recommended Flow:

/scan              → Map attack surface
    ↓
/specialist        → Deep-dive all domains (this skill)
    ↓
/audit             → Core vulnerability analysis
    ↓
/exploit           → Verify findings
    ↓
/report            → Generate final report

Or simply:

/start             → Runs everything automatically
```

## Quick Reference

| Command | What It Does |
|---------|--------------|
| `/specialist` | All 8 specialists in parallel |
| `/start` | Full assessment (includes specialists) |
| `/api` | API security only |
| `/injection` | Injection analysis only |
| `/crypto` | Cryptography only |
| `/supply-chain` | Dependencies only |
| `/file` | File security only |
| `/logic` | Business logic only |
| `/client` | Client-side only |
| `/config` | Configuration only |

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [kaivyy](https://github.com/kaivyy)
- **Source:** [kaivyy/perseus](https://github.com/kaivyy/perseus)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** yes
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-kaivyy-perseus-all
- Seller: https://agentstack.voostack.com/s/kaivyy
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.