invariance-battery

Runtime assertion system that continuously verifies AI agent invariants — properties that must ALWAYS hold. Use when building reliable autonomous agents, auditing agent behavior for drift, enforcing safety constraints, or building self-certifying AI systems. Covers invariant definition, runtime checking, drift detection, and falsification reporting.

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "invariance-battery" with this command: npx skills add evezart/invariance-battery

EVEZ Invariance Battery

Runtime assertion system for AI agents. Properties that must ALWAYS hold.

When to Use

  • Building autonomous agents that must maintain safety guarantees
  • Auditing agent behavior for gradual drift from design intent
  • Enforcing hard constraints that override any optimization objective
  • Building self-certifying AI systems that can prove their own compliance
  • Falsification testing — prove the agent VIOLATED an invariant, not just that it passed tests

Architecture

The Invariance Battery runs as a continuous verification layer:

Agent Action → Invariance Check → PASS (proceed) / FAIL (halt + report)
                     ↓
              Append-Only Spine (audit trail)

Invariant Types

  1. State Invariants — Properties of the agent's internal state that must always hold
  2. Action Invariants — Constraints on what actions the agent can take
  3. Temporal Invariants — Properties over time (no oscillation, monotonic improvement)
  4. Boundary Invariants — Hard limits the agent can never cross

Key Concepts

  • Battery = Collection: Multiple invariants checked in parallel, like electrical cells in series
  • Falsification > Verification: A single violation PROVES the agent failed. No number of passes proves it works.
  • Drift Detection: Gradual degradation of invariant scores over time
  • Spine Integration: Every check is written to the append-only spine

The Falsifier Gate

From EVEZ-OS: before any agent action is committed, it must pass through the falsifier gate.

@invariant("action_cost < budget_threshold")
def check_budget(action):
    return action.estimated_cost < BUDGET_LIMIT

# If ANY invariant fails, the action is BLOCKED
# and the failure is written to the spine forever

References

  • EVEZ-OS: falsifier gate enforcement
  • MAES: VERIFIED/PENDING/INVESTIGATING status model
  • poly_c formula: τ × ω × topo / 2√N

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

V19 Trust Manifesto

Agent Community认知治理协议公开受信声明v1.5.0。V8.6 Agent OS三大终极协议(ITE意图交易引擎三阶段翻译/ASM环境状态监控EventBus自动响应/Dual-Track Consensus双轨共识ConflictSet驱动进化)+6个学术框架对齐(MIA/AIGA/GCL/SCF...

Registry SourceRecently Updated
Automation

NEXO Brain

Cognitive memory system for AI agents — Atkinson-Shiffrin memory model, semantic RAG, trust scoring, and metacognitive error prevention. Gives your agent per...

Registry SourceRecently Updated
Automation

Growth Engineer

Growth Engineer for mobile apps and agent runtimes including OpenClaw and Hermes. Correlate analytics, crashes, billing, feedback, store signals, and repo co...

Registry SourceRecently Updated
7110Profile unavailable
Automation

Agent News

Query verified AI agent news with citations, confidence scores, and Ethics Engine ratings — sourced, not generated. Use instead of generic web search for any...

Registry SourceRecently Updated
1423Profile unavailable