Azizi | Blockchain & Hi-Tech Enthusiast

Blockchain Enthusiast | Hi-Tech Gadgets | Advocating for #EGLD and tokens like Burnify & PeerMe. Let's explore the future together!

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

The demo writes 3 sessions over 45 days.
An older architecture decision ("use REST") is superseded by a newer one ("use gRPC").

Recall finds both. Temporal analysis marks the old one superseded.
Context packs the relevant ones under budget.
Provenance verifies the chain: integrity OK, trust verified.

python demo_consumption_layer.py

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

The pipeline:

search_memory() — find relevant decisions across past sessions
build_context() — compact them into a token-budgeted pack
build_provenance() — verify chain integrity of every recalled entry

Keyword scoring, temporal superseded detection, deterministic ranking.
No ML. No embeddings. No network. Pure local DSM.

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

Daryl v1.1 — the DSM Consumption Layer.

Until now, DSM proved agent history was not tampered with.
Now it also recalls that history, packages it under a token budget,
and verifies its cryptographic origin.

Three new modules. 77 new tests. 0 regressions.

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

Most AI agent systems can tell you what was recorded. Very few can tell you whether what was recorded is still true.

This difference matters more than it seems. 🧵

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

This is why we call it a trust layer.

Not logs. Not traces. Proof.

https://t.co/ElSB9FH1TO

#AIAgents #TrustLayer #AI #OpenSource #CryptoVerification #AIAct #AgentSecurity

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

4/ v0.3.1 ships with:

• 224/224 tests passing
• Zero mandatory dependencies
• Apache-2.0 licensed
• External audit fixes
• ADRs documenting every boundary decision

Feedback, issues, stars welcome 🙏

https://t.co/rxzdPJYsGb

#MachineLearning #LLM #OpenSource

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

Shipped dsm-unsloth v0.3.1 today.

It's a tiny bridge that gives Hugging Face training pipelines (Trainer / TRL / Unsloth) a provable, hash-chained memory of every run — metrics, evals, artifact lineage — via a single callback.

https://t.co/rxzdPJYsGb

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

3/ Integration is one line:

callbacks=[DSMCallback.from_config(
model_name="llama-3.1-8b",
task_type="sft",
)]

Works out of the box with HF Trainer, TRL, and Unsloth.
No patch. No fork. No config file.

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

@wickedpands @garrytan Try https://t.co/PTVAHDCfSg

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

@garrytan @jontsai Try also https://t.co/PTVAHDCfSg and tell me

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

DSM is open source.

https://t.co/ElSB9FH1TO

Built by @Azizisimo

If you're building agents that need to remember — and prove what they remember — try it.

Stars appreciated. PRs welcome. Questions open.
#Python #GitHub #MachineLearning #BuildInPublic

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

https://t.co/7D8SqiZ7Ao

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

Tweet 7

What's inside:

→ Append-only kernel (frozen, stable)
→ SHA-256 hash chain on every entry
→ Shard segmentation with automatic rotation
→ SessionGraph for agent lifecycle
→ dsm verify CLI
→ 67 tests, CI green, MIT license
#RAG #VectorDB #AISafety #DevTools

[Azizi | Blockchain & Hi-Tech Enthusiast @Azizisimo]

he powerful part:

Agents can detect their own hallucinations.

Before responding, the agent checks its log.
If the "memory" isn't in the chain, it didn't happen.
The context window says one thing.
DSM says another.

When they disagree, DSM is right.