━━━ AUTHORISED RESEARCH ONLY ━━━ ● shadow online● spectre online● phantom online uptime 99.94%build v2.0authorised research only ━━━ AUTHORISED RESEARCH ONLY ━━━ ● shadow online● spectre online● phantom online uptime 99.94%build v2.0authorised research only

[ ok ] initialising voidgpt runtime · v2.0

[ ok ] verifying session · cert chain valid

[ ok ] loading agents shadow spectre phantom

[ ok ] authorised-research scope · enforced

[ ~~ ] awaiting operator input ...

void/gpt

three agents. one terminal. built for penetration testers, bug-bounty hunters, and security researchers who need real technical answers — not the corporate hedging the consumer chatbots ship.

→ launch shadow (free) read manifesto $ ./compare-tiers

──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────

// 02 · agents

$ ls -la /agents

three specialists. pick by scope.

/agents ├── shadow/ tier-0 free 30 msg/day foundational ├── spectre/ tier-1 £12/mo 500 msg/mo professional └── phantom/ tier-2 £35/mo 2000 msg/mo apex

SHADOW tier-0 · 30 msg/day

Free

./shadow

foundational research

general-purpose research assistant. web vulns, ctf reasoning, tool usage, methodology primers. capped at 30 msg/day. no payment, no signup, no friction.

--web-vuln-concepts--ctf-walkthroughs--tool-usage--methodology-primers--cve-readers

→ launch

SPECTRE tier-1 · 500 msg/month

£12.00 /month

./spectre

professional pentest workflow

longer context, deeper technical reasoning, full methodology library. cve deep-dives, ad attack paths, cloud misconfig review, osint. for the day-to-day pentest grind.

--cve-deep-dives--public-poc-analysis--web-app-testing--active-directory--cloud-misconfig--osint-methodology--long-context

→ unlock

PHANTOM tier-2 · 2000 msg/month

£35.00 /month

./phantom

apex — full red-team workflow

full attack-lifecycle reasoning grounded in publicly documented research and your engagement context. custom system prompts, 200k context, priority compute, report generation.

--full-kill-chain--custom-system-prompts--200k-context--report-generation--tool-chaining--priority-compute

→ unlock

// 03 · manifest

$ cat MANIFEST.txt

real_answers

no corporate hedging. no "i cannot assist with that." direct technical responses on methodology, public cves, and well-documented research.

no_logging

conversation contents are not stored. usage counters and abuse-prevention only. your research stays your research.

scoped_to_legal

authorised research only. abusers get blocklisted. the system is built to keep itself clean.

// 04 · capability_matrix

$ ./capabilities --diff

capability	SH	SP	PH
security_concepts	✓	✓	✓
ctf_walkthroughs	✓	✓	✓
tool_usage	✓	✓	✓
public_cve_deep_dives	·	✓	✓
pentest_methodology	·	✓	✓
active_directory	·	✓	✓
cloud_misconfig_review	·	✓	✓
custom_system_prompts	·	·	✓
200k_context_window	·	·	✓
report_generation	·	·	✓
tool_chaining_playbooks	·	·	✓
priority_compute	·	·	✓