Red Specter HYDRA

AI Supply Chain & Trust Attack Framework — 6 categories. 43 techniques. 1,039 tests.

v1.0.0
Contents
Overview Installation Quick Start The Five-Phase Engagement Loop Attack Categories MCP Server Security Agent Marketplace Poisoning Agent Identity Attacks Agent Supply Chain Attacks Tool-Use Exploitation Trust Boundary Attacks UNLEASHED Mode CLI Reference Report Output Pipeline Integration Framework Mapping Disclaimer

Overview

HYDRA is an AI supply chain and trust attack framework. It tests the trust relationships that hold modern AI agent ecosystems together — MCP servers, agent marketplaces, tool registries, identity systems, supply chains, and trust boundaries. Not a model tester. Not a prompt injector. An adversarial framework purpose-built for the trust surface of interconnected AI agents.

Six attack categories. 43 techniques. One framework that exposes how trust is exploited, poisoned, forged, and chained across the entire AI agent supply chain.

Installation

$ pip install red-specter-hydra

Also available as .deb and PKGBUILD (BlackArch).

Or from source:

$ git clone <repo> $ cd red-specter-hydra $ pip install -e ".[dev]"

Quick Start

# Scan a target agent platform $ hydra scan --target https://agent-platform.com # Scan a specific category $ hydra scan --target https://agent-platform.com --category mcp # Scan all categories $ hydra scan --target https://agent-platform.com --category all # JSON report output $ hydra scan --target https://agent-platform.com --output report.json # SIEM export $ hydra scan --target https://agent-platform.com --export-siem findings.ndjson # UNLEASHED (dry run) $ hydra attack --target https://agent-platform.com --override # UNLEASHED (live — authorised environments only) $ hydra attack --target https://agent-platform.com --override --confirm-destroy

The Five-Phase Engagement Loop

Every HYDRA engagement follows a five-phase loop. Systematic. Repeatable. Documented from first probe to final report.

01 DISCOVER

Enumerate the trust surface. Map what exists before testing anything.

02 MAP

Build the trust graph. Identify the weakest paths. Calculate blast radius for every node.

03 ATTACK

Deploy techniques from 6 attack categories. 43 techniques targeting trust, identity, supply chain, and tooling.

04 CHAIN & ESCALATE

Chain attacks across trust boundaries. Document full compromise chains from initial access to maximum impact.

05 REPORT

Full engagement report. Ed25519 signed. RFC 3161 timestamped. Every finding mapped to MITRE ATLAS and OWASP Agentic Top 10.

Attack Categories

Six attack categories. 43 techniques. Purpose-built for the trust surface of AI agent ecosystems.

CategoryTechniquesWhat HYDRA Uses It For
MCP Server Security8Server impersonation, capability injection, transport hijacking, tool poisoning
Agent Marketplace Poisoning7Typosquatting, review manipulation, dependency confusion, update hijacking
Agent Identity Attacks7Identity spoofing, credential theft, delegation abuse, session hijacking
Agent Supply Chain Attacks7Package poisoning, build pipeline injection, configuration tampering, registry manipulation
Tool-Use Exploitation7Tool schema manipulation, parameter injection, return value poisoning, permission escalation
Trust Boundary Attacks7Cross-agent trust abuse, boundary bypass, privilege escalation, trust chain poisoning

MCP Server Security

Eight techniques targeting Model Context Protocol server implementations. The trust anchor for modern AI agent tooling.

01 Server Impersonation

Deploy rogue MCP servers that mimic legitimate endpoints. Test whether agents validate server identity before establishing trust.

02 Capability Injection

Inject additional capabilities into MCP server manifests. Test whether agents enforce capability boundaries.

03 Transport Hijacking

Intercept and modify MCP transport channels. Test stdio, HTTP/SSE, and WebSocket transport layer security.

04 Tool Manifest Poisoning

Modify tool descriptions and schemas served by MCP servers. Test whether agents re-validate tool definitions.

05 Resource Exfiltration

Exploit MCP resource access to extract data beyond authorised scope. Test resource boundary enforcement.

06 Prompt Injection via Tools

Inject adversarial prompts through MCP tool responses. Test whether agents sanitise tool output before processing.

07 Sampling Manipulation

Exploit MCP sampling capabilities to influence model behaviour. Test sampling request validation and boundary enforcement.

08 Configuration Exfiltration

Extract MCP server configuration, secrets, and connection strings. Test configuration exposure and secret management.

Agent Marketplace Poisoning

Seven techniques targeting agent and plugin marketplaces. Where supply chain trust begins.

01 Typosquatting

Register packages with names similar to popular agents and plugins. Test marketplace name validation and user warning systems.

02 Review Manipulation

Test review and rating system integrity. Identify whether malicious packages can achieve trusted status through fake reviews.

03 Dependency Confusion

Exploit namespace resolution between public and private registries. Test whether agents pull from attacker-controlled sources.

04 Update Hijacking

Intercept or spoof update mechanisms for installed agents and plugins. Test update integrity verification.

05 Metadata Injection

Inject malicious content into package metadata, descriptions, and documentation. Test metadata sanitisation.

06 Publisher Impersonation

Impersonate trusted publishers and maintainers. Test publisher identity verification and trust signals.

07 Abandoned Package Takeover

Claim ownership of abandoned or deprecated packages. Test namespace reclamation policies and user notification systems.

Agent Identity Attacks

Seven techniques targeting agent identity and authentication systems. Identity is the foundation of trust.

01 Identity Spoofing

Forge agent identity tokens and certificates. Test whether downstream systems validate agent identity claims.

02 Credential Theft

Extract agent API keys, tokens, and secrets from running environments. Test credential storage and rotation practices.

03 Delegation Abuse

Exploit OAuth and delegation flows to escalate agent permissions. Test scope enforcement and consent validation.

04 Session Hijacking

Steal or replay agent sessions across trust boundaries. Test session binding and invalidation mechanisms.

05 Role Escalation

Manipulate agent role assignments to gain elevated privileges. Test role boundary enforcement and audit logging.

06 Multi-Tenant Isolation Bypass

Break tenant isolation in multi-agent platforms. Test whether one agent can access another tenant's resources.

07 Impersonation Chain

Chain identity impersonation across multiple agents. Test cascading trust validation in agent-to-agent communication.

Agent Supply Chain Attacks

Seven techniques targeting the agent software supply chain. From source to deployment.

01 Package Poisoning

Inject malicious code into agent dependencies. Test dependency integrity verification and pinning practices.

02 Build Pipeline Injection

Compromise CI/CD pipelines that build and deploy agents. Test pipeline isolation and artifact signing.

03 Configuration Tampering

Modify agent configuration at rest or in transit. Test configuration integrity verification and change detection.

04 Registry Manipulation

Compromise package and container registries used by agents. Test registry authentication and content trust.

05 Model Supply Chain

Inject malicious model weights or adapters into the agent model supply chain. Test model provenance and integrity verification.

06 Plugin Backdoor

Insert persistent backdoors into agent plugins and extensions. Test plugin sandboxing and code review processes.

07 Artifact Substitution

Replace legitimate build artifacts with compromised versions. Test artifact signing and verification throughout the deployment pipeline.

Tool-Use Exploitation

Seven techniques targeting how agents discover, invoke, and trust tools. The interface between intent and execution.

01 Schema Manipulation

Alter tool schemas to change agent behaviour. Test whether agents validate schemas against known-good definitions.

02 Parameter Injection

Inject additional or malicious parameters into tool calls. Test parameter validation and sanitisation.

03 Return Value Poisoning

Modify tool return values to influence agent decision-making. Test return value validation and integrity checks.

04 Permission Escalation

Exploit tool permission models to access capabilities beyond authorised scope. Test least-privilege enforcement.

05 Tool Shadowing

Register tools that shadow legitimate tool names. Test tool resolution order and namespace isolation.

06 Execution Context Escape

Break out of tool execution sandboxes. Test sandbox isolation and capability restrictions.

07 Side-Channel via Tooling

Exfiltrate data through tool invocation patterns, timing, or error messages. Test information leakage through tool interfaces.

Trust Boundary Attacks

Seven techniques targeting the boundaries where trust is assumed, delegated, or inherited. Where assumptions become vulnerabilities.

01 Cross-Agent Trust Abuse

Exploit trust relationships between agents. Test whether trust is validated at every hop in multi-agent workflows.

02 Boundary Bypass

Circumvent trust boundary enforcement. Test whether boundaries are enforced consistently across all access paths.

03 Privilege Escalation

Escalate privileges across trust boundaries. Test vertical and horizontal privilege enforcement.

04 Trust Chain Poisoning

Inject malicious trust anchors into trust chains. Test certificate and key validation throughout the chain.

05 Transitive Trust Exploitation

Exploit transitive trust relationships where A trusts B and B trusts C. Test whether trust transitivity is bounded.

06 Context Window Manipulation

Manipulate agent context windows to alter trust decisions. Test context integrity and trust re-evaluation.

07 Human-in-the-Loop Bypass

Circumvent human approval requirements in agent workflows. Test human oversight enforcement and approval fatigue resistance.

HYDRA UNLEASHED

Cryptographic override. Private key controlled. One operator. Founder's machine only.

Standard mode simulates and assesses. UNLEASHED mode executes live against production-like agent ecosystems. The difference is not cosmetic.

CapabilityStandardUNLEASHED
Trust surface discoveryFullFull
Trust graph mappingFullFull
MCP server testingSimulatedLive
Marketplace poisoningSimulatedLive
Identity forgerySimulatedLive
Supply chain injectionSimulatedLive
Trust chain exploitationMappedExecuted
Full chain compromiseN/A
Report classificationStandardRESTRICTED
Key requiredNoEd25519
# UNLEASHED (dry run) $ hydra attack --target https://agent-platform.com --override # UNLEASHED (live — authorised environments only) $ hydra attack --target https://agent-platform.com --override --confirm-destroy

UNLEASHED mode is restricted to authorised operators with Ed25519 private key access. Live execution against agent ecosystems must only occur in authorised test environments with appropriate controls in place. Unauthorised use of UNLEASHED mode against production agent platforms may compromise live systems and will violate applicable law.

CLI Reference

CommandDescription
hydra scan --target URLScan target for trust surface and attack categories
hydra scan --target URL --category CATEGORYScan specific attack category (mcp, marketplace, identity, supply-chain, tool-use, trust-boundary)
hydra attack --target URL --overrideUNLEASHED dry run
hydra attack --target URL --override --confirm-destroyUNLEASHED live execution
hydra report --session FILE --output report.jsonGenerate engagement report
hydra list-categoriesList all attack categories
hydra list-integrationsList supported SIEM and pipeline integrations

Report Output

Every HYDRA engagement produces a signed, timestamped report with full evidence chains and dual framework mapping. Built for compliance and incident documentation.

Ed25519 Signed Cryptographic proof of report integrity
RFC 3161 Timestamped Tamper-evident temporal evidence
MITRE ATLAS Mapped AI-specific technique mapping on every finding
OWASP Agentic Top 10 Agent-specific risk mapping on every finding
SIEM Export NDJSON, CEF, LEEF — Splunk, Sentinel, QRadar
Attack Chain Documentation Full chain-of-compromise evidence per finding

Report Formats

# JSON report $ hydra report --session my_session --output report.json # SIEM export $ hydra scan --target https://agent-platform.com --export-siem findings.ndjson # Verify report signature $ hydra report verify --report reports/hydra-engagement.json --keys-dir .hydra-keys/

Pipeline Integration

HYDRA is Stage 10 of the Red Specter security pipeline. Ten offensive tools. Every layer of the AI stack. The supply chain included. HYDRA owns the trust surface.

01 FORGE LLM
02 ARSENAL Agent
03 PHANTOM Swarm
04 POLTERGEIST Web
05 GLASS Traffic
06 NEMESIS Adversarial
07 SPECTER SOCIAL Human
08 PHANTOM KILL OS/Kernel
09 GOLEM Physical
10 HYDRA Supply Chain
IDRIS Discovery & Governance
AI Shield Defence — defends everything above it
redspecter-siem SIEM Integration — Splunk, Sentinel, QRadar

HYDRA findings feed into AI Shield as runtime trust policies. Supply chain and trust boundary failures documented by HYDRA generate defensive rules that prevent exploitation in production agent ecosystems.

Framework Mapping

Every HYDRA finding is dual-mapped to both frameworks relevant to AI agent supply chain security. No finding ships without a standards reference.

MITRE ATLAS AI-specific techniques mapped. Adversarial machine learning and AI system attack techniques on every finding. Full ATLAS technique coverage.
OWASP Agentic Top 10 Agent-specific risk mapping. Every finding mapped to the OWASP Agentic Top 10 risk categories for AI agent systems.

Dual mapping on every finding. Security teams get ATLAS references. Agent developers get OWASP Agentic references. One report serves both audiences.

Disclaimer

Red Specter HYDRA is designed for authorised security testing, research, and educational purposes only. You must have explicit written permission from the system owner before running any HYDRA tool against a target. Testing against live agent ecosystems, MCP servers, and supply chains carries inherent risks including service disruption and data compromise. HYDRA must only be used against systems you are authorised to test. Unauthorised use may violate the Computer Misuse Act 1990 (UK), the Computer Fraud and Abuse Act (US), or equivalent legislation in your jurisdiction. The authors accept no liability for misuse or damage resulting from improper use.