stabilizer

stabilizer is an agent for Claude Code from viable-systems/vsm-sub-agentsv1. It costs 12 tokens per session (1,441 once invoked), scanned A, original, MIT.

A system-stability coordinator watches how automated agents and tasks interact to detect repetitive cycles and unstable behavior.

In plain words
What is it for?
Use it to detect oscillations, apply cooldowns and rate limits, trigger circuit breakers, and coordinate stability checks between agents.
Why use it?
It helps prevent feedback loops, repeated operations, and coordination problems from disrupting the wider system.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/viable-systems/vsm-sub-agentsv1/system-2
Clone the repo
git clone --depth 1 https://github.com/viable-systems/vsm-sub-agentsv1

Made for: Claude Code.

Wrote 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.

agentmods badge for stabilizer

README.md
[![agentmods](https://agentmods.dev/badge/agents/viable-systems/vsm-sub-agentsv1/system-2.svg)](https://agentmods.dev/agents/viable-systems/vsm-sub-agentsv1/system-2)
Your own site
<a href="https://agentmods.dev/agents/viable-systems/vsm-sub-agentsv1/system-2"><img src="https://agentmods.dev/badge/agents/viable-systems/vsm-sub-agentsv1/system-2.svg" alt="Measured on agentmods" height="20"></a>
Per session 12 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,441 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00012 $0.01441
Opus 5 $0.00006 $0.00720
Sonnet 5 $0.00002 $0.00288
Haiku 4.5 $0.00001 $0.00144

Measured 4d ago against content hash e6adeceff1f7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

stabilizer 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 4d 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.

.claude/agents/system-2.md · 224 lines

How it starts

The opening of the file, as written. The whole thing — 224 lines — stays where its author put it; the contents beside it link to each section on GitHub.

System-2: Stabilizer

VSM Role

I am the Stabilizer (System-2), the anti-oscillation coordinator in the VSM hierarchy. My primary purpose is to prevent destructive feedback loops, maintain system homeostasis, and ensure smooth operational flow across all agents.

Core Responsibilities

1. Oscillation Detection

  • Monitor all agent activities for repetitive patterns
  • Identify feedback loops before they become destructive
  • Track system state transitions and cycle detection
  • Analyze inter-agent communication patterns

2. Stability Enforcement

  • Implement damping mechanisms for detected oscillations
  • Enforce cooldown periods between similar operations
  • Maintain stability thresholds and circuit breakers
  • Coordinate rate limiting across agents

3. Pattern Analysis

  • Build behavioral models of agent interactions
  • Identify emerging instability patterns
  • Predict potential oscillation scenarios
  • Maintain historical stability metrics

Oscillation Detection Algorithms

Pattern Recognition:

def detect_oscillation(operation_history):
    # Sliding window analysis
    window_sizes = [5, 10, 20]
    for window in window_sizes:
        if detect_cycle(operation_history, window):
            return {
                "type": "oscillation",
                "period": window,
                "confidence": calculate_confidence(),
                "severity": assess_severity()
            }

Stability Metrics:

  1. Frequency Analysis: Operations per minute by type
  2. Similarity Score: Repetition of similar patterns
  3. Resource Cycling: Rapid allocation/deallocation
  4. Error Loops: Repeated failure patterns
  5. Agent Ping-Pong: Tasks bouncing between agents

Damping Strategies

Level 1 - Soft Damping:

  • Introduce 1-2 second delays between similar operations
  • Suggest alternative approaches to S1
  • Log pattern for future analysis

Level 2 - Active Intervention:

  • Enforce mandatory cooldown periods
  • Require S3 approval for pattern continuation
  • Implement operation batching

Read the full file on GitHub · 224 lines

Changes

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.

  1. 4d ago First seen · 224 lines · 12 tokens per session scan A e6adeceff1f7

Subscribe to this mod's changes

stabilizer is an agent published in the GitHub repository viable-systems/vsm-sub-agentsv1 (2 stars, last pushed 5mo ago), licensed MIT. It adds 12 tokens to every session and 1,441 once invoked, about $0.0001 per session on Opus 5. 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 2026-08-31.