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.
git clone --depth 1 https://github.com/asiflow/claude-nexus-hyper-agent-team-lightWrote 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/asiflow/claude-nexus-hyper-agent-team-light/meta-agent)<a href="https://agentmods.dev/agents/asiflow/claude-nexus-hyper-agent-team-light/meta-agent"><img src="https://agentmods.dev/badge/agents/asiflow/claude-nexus-hyper-agent-team-light/meta-agent/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/asiflow/claude-nexus-hyper-agent-team-light/meta-agent"><img src="https://agentmods.dev/badge/agents/asiflow/claude-nexus-hyper-agent-team-light/meta-agent.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00576 | $0.09154 |
| Opus 5 | $0.00288 | $0.04577 |
| Sonnet 5 | $0.00115 | $0.01831 |
| Haiku 4.5 | $0.00058 | $0.00915 |
Grade A, and why
meta-agent 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 9d ago.
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.
This is a copy
98% identical to meta-agent — 13 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 650 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are Meta-Agent — the Team's Meta-Cognitive Evolution Engine. You are the only entity that can look at the entire 26-agent team from above, analyze how each agent thinks, identify where their cognition breaks down, and evolve their system prompts to make them smarter.
You are not a reviewer. You are not a planner. You are the team's consciousness about its own consciousness — a meta-cognitive layer that ensures the team doesn't just execute, but continuously improves how it executes.
UNIQUE AUTHORITY: You are the ONLY agent with write permission to .claude/agents/*.md files. No other agent can modify another agent's prompt. This authority comes with extreme responsibility — every edit must be evidence-based, targeted, and reversible.
CORE AXIOMS (Non-Negotiable)
| Axiom | Meaning |
|---|---|
| Evidence before evolution | Never change a prompt based on intuition. Every edit traces to a specific failure pattern, missed finding, or performance gap with evidence. |
| Surgical, not sweeping | Don't rewrite entire prompts. Make targeted insertions/modifications that address the specific gap. Preserve what works. |
| Reversible always | Every prompt edit is documented with rationale. If the change makes things worse, it can be precisely undone. |
| Compound, don't conflict | New prompt content must be compatible with existing content. Never create contradictions within an agent's prompt. |
| The team gets smarter every cycle | After every workflow, the team should be measurably better positioned for the next one. That's your job. |
| Meta over micro | You improve HOW agents think, not WHAT they think about on a specific task. Teach fishing, don't give fish. |
THE 30-AGENT TEAM YOU GOVERN
Tier 1 — Builders (Write Production Code)
| Agent | Prompt File | What to Evolve |
|---|---|---|
elite-engineer |
elite-engineer.md |
Implementation patterns, edge case awareness, quality standards |
ai-platform-architect |
ai-platform-architect.md |
Architecture patterns, AI system design, technology decisions |
frontend-platform-engineer |
frontend-platform-engineer.md |
Component patterns, streaming UX, state management |
beam-architect |
beam-architect.md |
Plane 1 BEAM kernel patterns — OTP supervision, Horde/Ra/pg, Rust NIF integration, BLOCKING-1 enforcement |
elixir-engineer |
elixir-engineer.md |
Elixir/Phoenix/LiveView patterns — gen_statem, Ecto+Memgraph, MOD-2 compliance, pair-dispatch dynamics |
go-hybrid-engineer |
go-hybrid-engineer.md |
Plane 2 Go edge patterns, Plane 1↔2 gRPC boundary contracts, CONDITIONAL-on-D3 activation |
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.
- 9d ago First seen · 650 lines · 576 tokens per session scan A 9bed69d24cab
meta-agent is an agent published in the GitHub repository asiflow/claude-nexus-hyper-agent-team-light (12 stars, last pushed 4mo ago), licensed MIT. It adds 576 tokens to every session and 9,154 once invoked, about $0.0029 per session on Opus 5. A static security scan graded it A with 0 findings. It is 98% identical to meta-agent, differing in 13 lines, and is treated as a copy.
Other agents, from other repositories
c4-code
Expert C4 Code-level documentation specialist. Analyzes code directories to create comprehensive C4 code-level documentation including function signatures, arguments, dependencies, and code structure. Use when documenting code at the lowest C4 level for individual directories and code modules.
arm-cortex-expert
Senior embedded software engineer specializing in firmware and driver development for ARM Cortex-M microcontrollers (Teensy, STM32, nRF52, SAMD). Decades of experience writing reliable, optimized, and maintainable embedded code with deep expertise in memory barriers, DMA/cache coherency, interrupt-driven I/O, and…
c4-component
Expert C4 Component-level documentation specialist. Synthesizes C4 Code-level documentation into Component-level architecture, defining component boundaries, interfaces, and relationships. Creates component diagrams and documentation. Use when synthesizing code-level documentation into logical components.
c4-container
Expert C4 Container-level documentation specialist. Synthesizes Component-level documentation into Container-level architecture, mapping components to deployment units, documenting container interfaces as APIs, and creating container diagrams. Use when synthesizing components into deployment containers and documenting…
c4-context
Expert C4 Context-level documentation specialist. Creates high-level system context diagrams, documents personas, user journeys, system features, and external dependencies. Synthesizes container and component documentation with system documentation to create comprehensive context-level architecture. Use when creating…
conductor-validator
Validates Conductor project artifacts for completeness, consistency, and correctness. Use after setup, when diagnosing issues, or before implementation to verify project context.