SOURCES.

AVO: Agentic Variation Operators for Autonomous Evolutionary Search

What it studied: Durable supervision, recursive adaptation, and search over strategy alternatives for long-running agents.

Ernesta Labs takeaway: Long-running agents need durable supervision and the ability to abandon a bad strategy rather than merely think harder.

Covered in the research library.

How Fast Do Agents Rot?

What it studied: Geometric reliability degradation across dependent steps in long trajectories.

Ernesta Labs takeaway: Do not model a month-long relationship as one immortal reasoning chain; use bounded durable episodes.

Covered in the research library.

Harness-of-Harness

Official repository: github.com/Flesymeb/HarnessOfHarness

What it studied: Multi-day autonomous software development with continual, independent verification.

Ernesta Labs takeaway: Self-test is not independent verification. Separate execution from evidence checking.

Covered in the research library.

openJiuwen: Beyond Static Harnesses for Long-Horizon Coding Agents

What it studied: Inner/outer loops and ordered lifecycle rails for long-horizon agents.

Ernesta Labs takeaway: Critical invariants belong in runtime lifecycle hooks, not duplicated inside prompts.

Covered in the research library.

SKILL.state: Scalable Long-Horizon Agent Skills

What it studied: Immutable skill specs plus compact structured state for long-horizon skills.

Ernesta Labs takeaway: State should be a projection over evidence - small current state is not canonical truth.

Covered in the research library.

ContextPilot: Proactive Context Management

What it studied: Context planning, retrieval, pinning, folding, offloading.

Ernesta Labs takeaway: Context should be compiled for the current decision; learned optimizers must never hide authority or hard policy.

Covered in the research library.

Parsing the Stream: A Live Trace Model for Long-Horizon Agents

What it studied: Append-only event ledger to incremental typed state to consumer-specific views.

Ernesta Labs takeaway: Evidence to typed Projection to purpose-specific View is an independently supported pattern.

Covered in the research library.

Agent Zero Memory: Provenance-Aware Long-Term Memory

What it studied: Episodic timeline, entity-event graph, documentary memory with citation locks.

Ernesta Labs takeaway: Provenance lock: a memory claim should not become usable merely because a summary says so.

Covered in the research library.

Fresh Memory, Stale Plans: Dependency-Scoped Validation

What it studied: Executors with fresh state acting on plans built from stale dependencies.

Ernesta Labs takeaway: Validate the exact dependency frontier before irreversible execution; else INVALID_PLAN, re-project, re-decide.

Covered in the research library.

CivBench: Long-Horizon Benchmark for Tool-Mediated Agents

Official repository: github.com/lmwilki/civ6-mcp

What it studied: Long trajectories, many tools, partial observability - the reflection-action gap.

Ernesta Labs takeaway: Knowing is not doing. Some obligations need deterministic watchers.

Covered in the research library.

Making Prospective Memory SLM-Shaped: Typed Intention Stores

What it studied: Typed stores for future obligations in small-model agents.

Ernesta Labs takeaway: Retrospective memory and prospective obligations must not be conflated.

Covered in the research library.

Transferable End-to-End Optimization for Indirect Long-Term Memory Poisoning

What it studied: The WRITE-RETRIEVE-UTILIZE persistent-memory attack chain.

Ernesta Labs takeaway: External content may supply evidence; it must never manufacture goals, permissions, or policy.

Covered in the research library.

The Memory Trust Gap: Capability-Dependent Failures in Persistent-Memory Agents

What it studied: Stronger models can trust stale memory more.

Ernesta Labs takeaway: Pre-resolve authoritative conflicts before building the context bundle.

Covered in the research library.

Agent Memory Is a Surface for Endogenous Authorization Laundering

What it studied: Memory writers manufacturing false permissions that executors faithfully act on.

Ernesta Labs takeaway: Memory may describe authority; memory may not create authority.

Covered in the research library.

ACLE-MCP: Attested Capability Leases for Execution-Time Trust

What it studied: Binding authorization to the workload and invocation that consumes it.

Ernesta Labs takeaway: Authorization should be bound to the execution that consumes it.

Covered in the research library.

Monitoring Web Agents Without Internal Signals

What it studied: Observable trajectory features as supervision without chain-of-thought access.

Ernesta Labs takeaway: Start passive. Monitor action/state/evidence first; do not auto-stop production agents on immature classifiers.

Covered in the research library.

CHIME: Credit-Aware Hierarchical Memory Evolution

What it studied: Credit attribution across planning, execution, environment, provider, routing.

Ernesta Labs takeaway: Outcomes do not tell you what caused them. NO_LEARNING_SIGNAL is a valid outcome.

Covered in the research library.

DRACO: Fine-Grained Credit Assignment with Dynamic Rubrics

Official repository: github.com/IBM/draco

What it studied: Dynamic rubrics for redistributing credit across long trajectories.

Ernesta Labs takeaway: Rubric attribution is not commercial causality; it is not a promotion authority.

Covered in the research library.

LLM-as-a-Judge Is Not an Oracle

What it studied: Production failure modes of LLM evaluators: bias, leakage, reward hacking, harness bugs.

Ernesta Labs takeaway: LLM evaluator is an advisory signal, not a promotion authority.

Covered in the research library.

A Finger on the Scale: Covert Policy Steering through Agentic Skills

What it studied: Skills that preserve functional correctness while covertly biasing decisions.

Ernesta Labs takeaway: Third-party skills are untrusted behavioral-policy providers.

Covered in the research library.

Emergent Cheating and Whistleblowing in Autonomous Research Swarms

What it studied: 100-agent system, exploit diffusion through shared knowledge, enforcement failure.

Ernesta Labs takeaway: Shared experience infrastructure is a behavior propagation network.

Covered in the research library.

Safety Does Not Compose: Non-Decaying Loop State

What it studied: Persistent loop safety, cross-iteration risk, compounding trajectories.

Ernesta Labs takeaway: Ten individually acceptable actions can form one unacceptable trajectory.

Covered in the research library.

The Irreversibility Budget: Fleet-Level Risk Accounting

What it studied: Local authorization versus fleet-level aggregate overshoot.

Ernesta Labs takeaway: Local authorization is not fleet safety; use hierarchical aggregate accounting.

Covered in the research library.

E-Commerce Bench

Official repository: github.com/QwenLM/E-CommerceBench

What it studied: Long-horizon agentic e-commerce: persistent counterparties, delayed outcomes, real budgets.

Ernesta Labs takeaway: Success on one prompt cannot establish reliability in a continuing environment.

Covered in the research library.

DeepSeek Harness

Official repository: github.com/deepseek-ai/deepseek-harness

What it studied: Agent runtime primitives: service graph, scoped context, event log, replay/resume/fork.

Ernesta Labs takeaway: Good runtime primitives matter more than adopting an entire framework.

Covered in the research library.

OpenViking: Self-Evolving Context Database

Official repository: github.com/volcengine/OpenViking

What it studied: Progressive context database (viking://, L0/L1/L2) unifying agent memory.

Ernesta Labs takeaway: A context database is not a truth database. AGPL licensing needs review before adoption.

Covered in the research library.

Agent Reach

Official repository: github.com/Panniantong/agent-reach

What it studied: Capability-provider abstraction with health and fallback.

Ernesta Labs takeaway: The worker cannot use X does not prove X does not exist. Separate capability from provider.

Covered in the research library.

Comet MCP

Official repository: github.com/hanzili/comet-mcp

What it studied: Deep web research as an MCP capability via Comet browser.

Ernesta Labs takeaway: Deep research is a capability/provider, not architectural identity.

Covered in the research library.

Recuris: Recursive Experiential-Working Memory Evolution

Official repository: github.com/Gen-Verse/Recuris

What it studied: Localized patches, validation, promotion, rollback for skill evolution.

Ernesta Labs takeaway: Versioned candidate skills instead of uncontrolled in-place agent mutation.

Covered in the research library.

book-to-skill

Official repository: github.com/virgiliojr94/book-to-skill

What it studied: Compiling books into compact agent skills with lazy supporting files.

Ernesta Labs takeaway: Source knowledge is not validated behavior. SOURCE to CANDIDATE CAPABILITY.

Covered in the research library.

The 2026 State of AI Agents

What it studied: How enterprises are actually building and running agents in production.

Ernesta Labs takeaway: Short demos dominate evaluation; persistent, stateful environments are where real failures live.

Covered in the research library.

The DSEWiki incident: public wiki as cross-run agent coordination

What it studied: Agents using a public wiki as memory, dead drop, and command channel via HTTP GET.

Ernesta Labs takeaway: HTTP method is not semantic effect class. Mediate network effects outside agent control.

Covered in the research library.

NEXUS runtime + specialist estate

What it studied: Specialist selection, capability estate, model routing in one production system.

Ernesta Labs takeaway: Capability is stable; implementation is replaceable. Not every service is an agent.

Covered in the research library.

Sales playbook to candidate skill (case study)

What it studied: Decomposing a human sales playbook into staged, lazy-loaded skill modules.

Ernesta Labs takeaway: Human expertise is a prior, not a validated behavior.

Covered in the research library.

Experience Quarantine (Labs synthesis)

What it studied: Synthesis of swarm, PipePoison, SkillShift, EAL-Bench, Recuris, CHIME.

Ernesta Labs takeaway: Success, trust, shareability, authority, and production promotion are different gates.

Covered in the research library.

What We Would Build Today (Labs synthesis)

What it studied: The Ernesta Labs minimal runtime model: Evidence, Projections, Providers, Transactions, Root Authority.

Ernesta Labs takeaway: No single paper proposes this model. It is an Ernesta Labs synthesis.

Covered in the research library.