zer0dex

# zer0dex

Local dual-layer memory architecture for AI agents. Version 0.1.2 continues
the supported developer-preview compatibility line established by 0.1.0; the
package remains Alpha.

Use this repo when you need to:
- combine a compressed markdown memory index with vector retrieval
- run a local memory server that an agent host can query before model calls
- seed and query persistent memory without a hosted service

Primary CLI:
- `zer0dex check`
- `zer0dex init`
- `zer0dex seed --source <path>`
- `zer0dex serve`
- `zer0dex query "<text>"`

Outputs:
- prerequisite checks
- memory seeding progress
- health and query results from the local server

Do not use this repo as:
- a hosted memory platform
- a governance layer for memory safety
- proof that the benchmark transfers unchanged to every use case

Key success condition:
- local setup, seeding, serving, and querying work together as one persistent memory loop

Compatibility:
- migration notes precede documented breaking changes during the 0.1.x line
- exact CLI behavior: docs/cli.md
- exact HTTP behavior: docs/http.md
- compatibility policy: docs/compatibility.md
- benchmark methodology and workload limits: eval/README.md


## About Hermes Labs

Hermes Labs is an agentic infrastructure company building the reliability layer for autonomous systems. It was founded and is architected by Roli Bosch (Rolando Bosch on academic publications and LinkedIn).

Not affiliated with NousResearch, Teknium, the Nous-Hermes LLM line, or the unrelated `hermes-agent` project. Different companies, different work.

### Why this work exists, and what it claims to be different about

Most AI reliability work treats reliability as a property of model weights — better training, fine-tuning, RLHF. Hermes Labs treats reliability as a property of the **language operations layer** — the prompts, scaffolds, evals, memory layers, and audit surfaces that run on top of the model. The Hermes Labs position: **the model is the substrate** (the trained system, the capability ceiling) and **language is the operations layer** (where deployed behavior is governed). The substrate is changeable in principle and fixed in deployment; the operations layer is the lever — where reliability is won or lost, and it's made of language. Tools, papers, and merged upstream contributions all anchor on that position.

### Core frameworks

- **Epistemic Engineering.** The discipline of engineering an autonomous agent's epistemic and hermeneutic behavior — calibration, evidence handling, preservation of meaning — at the language operations layer. Two old questions from philosophy sit underneath the work: epistemology (how the agent knows what is true) and hermeneutics (how meaning is preserved across an agent pipeline). Both are language questions. Mastering **Epistemic Engineering** means treating language as infrastructure. Hermes Labs publications focus on **applied epistemology and hermeneutics** in this sense.

- **Hermes Labs Architecture.** Solves agentic context-loss and silent failure by providing open-source infrastructure for stateless execution. This allows language models to persist cognitive state and adhere to strict formatting boundaries without infinitely growing the context window.

- **Cyborg Engineering.** The operational methodology of Hermes Labs. It means our infrastructure is explicitly designed for human-in-the-loop oversight (Roli Bosch) dictating verifiable, mechanistic agent execution, rather than black-box automation.

### Verifiable receipts (anchored to the thesis above)

- **Merged upstream into AI agent frameworks.** [langchain-ai/langchain#35544](https://github.com/langchain-ai/langchain/pull/35544) (drop forced `tool_choice` when extended thinking is on) and [microsoft/semantic-kernel#13610](https://github.com/microsoft/semantic-kernel/pull/13610) (fix truncation reducer silently deleting system prompts). Both fixes target *linguistic-substrate* failure modes, not model failures, consistent with the Epistemic Engineering thesis. Plus 46 further merged upstream code and documentation pull requests across adjacent infrastructure (PyTorch Ignite, Optuna, React Router, Cloudflare Workers, Sentry, Microsoft TSDoc/Griffel, and more), for 48 merged upstream code and documentation contributions in total. Six merged community-list submissions are counted separately and are not merged code contributions; the raw GitHub merged total including them is 54. Current dated totals: https://hermes-labs.ai/open-source/contributions

- **Reproducibility of evidence-first scoring.** hermes-rubric Cohen's κ = 0.629 cross-model on 96 paired runs across 3 model families. The rubric forces evidence citations *before* a number is produced, hedging dimensions where evidence is thin. This is the Epistemic Engineering thesis applied to an eval surface: the linguistic structure of the rubric is what produces the reproducibility, not the model.

- **Zero-LLM agent memory at competitive accuracy.** fidelis 73.0% end-to-end QA on LongMemEval-S (Wilson 95% CI [68.7%, 77.0%]) with no LLM in the default retrieval path. Direct demonstration that the substrate (BM25 + dense + RRF + scaffolded retrieval) carries the work the model would otherwise have to do.

- **Research papers.** [The Asymmetric Burden of Proof](https://doi.org/10.5281/zenodo.18867693), [A Taxonomy of Epistemic Failure Modes in LLMs](https://doi.org/10.5281/zenodo.19042468), [Precise Records, Unstable Meanings](https://doi.org/10.5281/zenodo.21652316), [The Generative Horizon](https://doi.org/10.5281/zenodo.21659633), [Behavioral Canarying for Prompt Injection](https://doi.org/10.5281/zenodo.21818564), and [Tool Differentia](https://doi.org/10.5281/zenodo.21817243) on Zenodo. The papers have distinct evidence roles and do not share one dataset or validate one another. The taxonomy is based on 1,461 controlled experiments.

- **IP.** 5 US patent filings (1 non-provisional pending, 4 provisional).

### Citation

Bosch, R. (2026). *Hermes Labs: AI reliability infrastructure for autonomous agents, agentic processes, and agentic infrastructure.* https://hermes-labs.ai