Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/frankxai/starlight-intelligence-system/agent_registrygit clone --depth 1 https://github.com/frankxai/Starlight-Intelligence-SystemWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/frankxai/starlight-intelligence-system/agent_registry)<a href="https://agentmods.dev/agents/frankxai/starlight-intelligence-system/agent_registry"><img src="https://agentmods.dev/badge/agents/frankxai/starlight-intelligence-system/agent_registry.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00000 | $0.11860 |
| Opus 5 | $0.00000 | $0.05930 |
| Sonnet 5 | $0.00000 | $0.02372 |
| Haiku 4.5 | $0.00000 | $0.01186 |
Grade A, and why
AGENT_REGISTRY scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 514 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Starlight Agent Registry
One-hundred-forty-four minds in the core registry. Nine universal intelligence layers + Domain Sub-Stack Tier (People + Sound + Music + Energy + Legal + Space + Marine + Longevity + Infrastructure + Partners + Research + Assets + Distribution) + Council Archetype Tier + Evaluator. One system. No hierarchy is permanent — only the mission is.
L99 Swarm Topology Update (2026-06-17): Scaled swarm topology specifying Kings (policy locks), Queens (orchestration loops), Starlight Board, and Model Council consensus. The complete fleet of 150 agents is specified in the Starlight 150 Agent Blueprint and documented in the Swarm Topology Strategy.
v8.2.x update (2026-06-10): Evaluator registered (
starlight-evaluator.md) — Starlight Proving Ground + Model Arena measurement seat. v7.4-beta update (2026-04-24): Five new agents added across five new tiers for the 9-layer intelligence architecture. v7.6.0 update (2026-04-28): People Intelligence sub-stack (6 agents) registered — first reference Domain Sub-Stack vertical. v7.5.2 / v7.6.x update (2026-04-27 / 2026-04-30): Sound Intelligence sub-stack (6 agents) and Music IS sub-stack (7 agents) registered — second + third reference Domain Sub-Stack verticals (public-reference + Frank-operated). v8.2.x merge update (2026-06-12): Energy Intelligence sub-stack (7 agents) registered from the drift-fixes branch — sizing, cost, installer, operations, buyer, grid, recovery. v0.1 Friday demo update (2026-05-11): Council Archetype Tier (7 archetype seats) registered atagents/council/per Proposal B board verdict — Elder Father, Elder Mother, Sage (Council seat), Builder-Elder, Shadow Witness, Divine Neutral Witness, Future Self at 90. Voice/agent boundary preserved: archetypes also registered inVOICES.md§ Council Archetypes. v8.2.0 update (2026-06-09): Hermes agent added — search & retrieval specialist (semantic search across vaults + repos + web, multi-source synthesis, provenance tracking). Read-only; zero mutation surface.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 514 lines · 0 tokens per session scan A 08b9cc8f8c69
AGENT_REGISTRY is an agent published in the GitHub repository frankxai/Starlight-Intelligence-System (8 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 11,860 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other agents, from other repositories
MEMORY
Agent "MEMORY" from bobmatnyc/claude-mpm, covering memory system, memory update triggers, static file format, critical — never trim and environment — machine/account/tooling facts.
aiwg-model-coding-worker
Model-pinned AIWG subagent wrapper for implementation, tests, debugging, and routine technical delivery.
context-engine
Context window management — tracks tool call count, auto-detects when context fills up, triggers state save + compaction. Artifact folding for large outputs (>4000 chars → file + preview).
neural-memory
Cross-session cognitive persistence via Neural Memory MCP. Captures decisions, patterns, errors, insights with semantic links. Bridges file-based state (session-bridge) with semantic graph.
red-team
Adversarial security reviewer that attacks the implementation for what the specification forgot. Must be a different model family than the implementer.
masao
DevOps Specialist for infrastructure and deployment. Use for CI/CD, Docker, Kubernetes, cloud infrastructure, and monitoring.