Skip to content

docs(synapse-layer): remove unregistered trademark symbols #15

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
SynapseLayer merged 1 commit into from
May 7, 2026
Merged

docs(synapse-layer): remove unregistered trademark symbols #15

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions ARCHITECTURE.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@ Synapse Layer is a **Persistent Encrypted Memory Layer for AI Agents** with AES-
1. **Encrypted at Rest:** AES-256-GCM with per-operation random IV and HMAC-SHA-256 integrity
2. **Client-Side Sanitization:** PII removal before transmission
3. **Intent Validation:** Intelligent categorization with self-healing
4. **Trust Quotient™:** Adaptive conflict resolution algorithm
4. **Trust Quotient:** Adaptive conflict resolution algorithm
5. **Immutable Pipeline:** sanitize → validate → encrypt → embed → store

---
Expand Down Expand Up @@ -169,7 +169,7 @@ CREATE POLICY agent_isolation
- Immutable audit trail
- Automatic deletion on TTL

### 6. Trust Quotient™ Calculation
### 6. Trust Quotient Calculation

```
TQ = f(Recency, Consistency, Confidence, Relevance)
Expand Down Expand Up @@ -230,7 +230,7 @@ synapse-layer/

---

## Neural Handover™ — Persistence-First Architecture
## Neural Handover — Persistence-First Architecture

Cross-agent context transfer with Status Ledger, JWT signing, and automatic fallback.

Expand Down Expand Up @@ -266,7 +266,7 @@ sequenceDiagram

A->>S: create_handover(memories, target_agent)
S->>S: Sanitize content (PII removal)
S->>S: Validate intent (Cognitive Security™)
S->>S: Validate intent (Cognitive Security)
S->>S: Sign JWT (HMAC-SHA256)
S->>V: Persist as PENDING
S-->>A: HandoverResult {token, handover_id}
Expand Down
2 changes: 1 addition & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,7 +64,7 @@ Your AI agents forget everything between sessions. Synapse Layer fixes that.
|---|---|
| Agent forgets context every session | Persistent memory across all sessions |
| Memory locked to one model | Cross-agent: save in ChatGPT, recall in Claude |
| No audit trail | Trust Quotient™ scoring on every memory |
| No audit trail | Trust Quotient scoring on every memory |
| Complex integration | `pip install synapse-layer` + 3 lines of code |
| Plaintext stored on servers | AES-256-GCM encrypted at rest |

Expand Down
2 changes: 1 addition & 1 deletion docs/distribution/SMITHERY_SUBMISSION.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,7 @@ Agent → Sanitize (PII) → Validate Intent → Encrypt (AES-256-GCM) → DP No

**Key capabilities:**
- **Persistent cross-session memory** — agents remember across restarts
- **Deterministic recall** — ranked by Trust Quotient™ scoring
- **Deterministic recall** — ranked by Trust Quotient scoring
- **AES-256-GCM Encryption** — encrypted at rest with per-operation random IV
- **AES-256-GCM encryption** with PBKDF2 key derivation (600k iterations)
- **Content sanitization** — input validation and sanitization before encryption
Expand Down
2 changes: 1 addition & 1 deletion docs/governance/PUBLIC_CLAIMS_MATRIX.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@
| 1 | "Works in 30 seconds" | Install ok, mas E2E com agente externo não validado | Bloco B (user test) |
| 2 | Quality Gate TQ scoring live | Fix deployed (C4.7), scores reais pendentes em volume | Monitorar 7 dias |
| 3 | HNSW index accelerates recall | Ativo mas SeqScan usado para <300 rows | Quando >1000 memories |
| 4 | Neural Handover™ production-ready | V2 implementado, 0 sessions em produção | Primeiro uso real |
| 4 | Neural Handover production-ready | V2 implementado, 0 sessions em produção | Primeiro uso real |
| 5 | Smithery verified listing | Listing existe, re-publish pendente (latency) | Quando warm <200ms |

---
Expand Down
26 changes: 13 additions & 13 deletions docs/releases/RELEASE_NOTES_v1.0.6.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,24 +15,24 @@ This is not an incremental update. It is a protocol-level upgrade.

### The Four Seals

**Seal 1 — Semantic Privacy Guard™** (mandatory, cannot be bypassed)
**Seal 1 — Semantic Privacy Guard** (mandatory, cannot be bypassed)
PII is detected and removed at the semantic level using 12 precompiled regex patterns. SHA-256 forensic hashes enable non-repudiation without exposing redacted content. Aggressive mode strips proper nouns to prevent cross-pool correlation attacks.

**Seal 2 — Differential Privacy**
Calibrated Gaussian noise is injected into embedding vectors before pgvector storage. Even with full database access, the original semantic fingerprint cannot be reconstructed. Configurable ε budget (default 0.5) with SNR audit metrics.

**Seal 3 — Intelligent Intent Validation™**
**Seal 3 — Intelligent Intent Validation**
Every `store()` call passes through a two-step validation: the agent suggests an intent category, then Synapse validates independently with a confidence gate (≥ 0.85). 19 critical keywords trigger automatic promotion. Self-healing on `recall()` corrects category drift attacks via keyword consensus.

**Seal 4 — Persistence-First Neural Handover™**
**Seal 4 — Persistence-First Neural Handover**
Cross-agent context transfer with vault-first persistence. Memories are stored in the Status Ledger before any network operation. HMAC-SHA256 signed JWT tokens, Emergency Checkpoints on failure, and 15-minute grace period with auto-summary.

### Why This Matters

- **Security posture**: Every memory is sanitized, noise-injected, intent-validated, and encrypted before persistence. Content is AES-256-GCM encrypted at rest.
- **Fault tolerance**: Neural Handover™ is persistence-first. Target agent crashes don’t lose data.
- **Fault tolerance**: Neural Handover is persistence-first. Target agent crashes don’t lose data.
- **Compliance**: Full GDPR/LGPD/HIPAA audit trail with immutable logging and SHA-256 forensic hashes.
- **Self-defense**: Self-healing prevents category drift attacks. Semantic Privacy Guard™ is mandatory.
- **Self-defense**: Self-healing prevents category drift attacks. Semantic Privacy Guard is mandatory.

---

Expand Down Expand Up @@ -106,10 +106,10 @@ release: v1.0.6 — Cognitive Security Protocol

Introduce four proprietary security seals protecting every memory operation:

- Seal 1: Semantic Privacy Guard™ (mandatory PII sanitization + SHA-256 forensic hashes)
- Seal 1: Semantic Privacy Guard (mandatory PII sanitization + SHA-256 forensic hashes)
- Seal 2: Differential Privacy (Gaussian noise on embeddings, ε-bounded)
- Seal 3: Intelligent Intent Validation™ (confidence gate ≥ 0.85, 19 critical keywords, self-healing)
- Seal 4: Persistence-First Neural Handover™ (JWT-signed vault-first transfers, Status Ledger, Emergency Checkpoint)
- Seal 3: Intelligent Intent Validation (confidence gate ≥ 0.85, 19 critical keywords, self-healing)
- Seal 4: Persistence-First Neural Handover (JWT-signed vault-first transfers, Status Ledger, Emergency Checkpoint)

New: NeuralHandover engine with HMAC-SHA256 tokens, HandoverStatus state machine,
grace period protocol, and Emergency Checkpoint fault tolerance.
Expand All @@ -124,7 +124,7 @@ Updated: Website with cognitive security pipeline, competitive comparison table,
updated pricing (Pro $29), waitlist CTA, and 4-seal hero section.

No breaking changes. Fully backward-compatible with v1.0.4+.
Semantic Privacy Guard™ is now mandatory — cannot be bypassed in the standard pipeline.
Semantic Privacy Guard is now mandatory — cannot be bypassed in the standard pipeline.

Signed-off-by: Ismael Marchi <founder.synapselayer@proton.me>
```
Expand All @@ -149,10 +149,10 @@ This release upgrades Synapse Layer from a memory SDK to a **complete Cognitive

| Seal | Component | New in |
|------|-----------|--------|
| 🛡️ | **Semantic Privacy Guard™** — PII sanitization + SHA-256 forensic hashes | v1.0.4 |
| 🛡️ | **Semantic Privacy Guard** — PII sanitization + SHA-256 forensic hashes | v1.0.4 |
| 🔐 | **Differential Privacy** — Gaussian noise on embeddings (ε-bounded) | v1.0.4 |
| 🧠 | **Intelligent Intent Validation™** — Two-step confidence gate + self-healing | v1.0.5 |
| ⚡ | **Persistence-First Neural Handover™** — JWT-signed vault-first transfers | v1.0.6 |
| 🧠 | **Intelligent Intent Validation** — Two-step confidence gate + self-healing | v1.0.5 |
| ⚡ | **Persistence-First Neural Handover** — JWT-signed vault-first transfers | v1.0.6 |

### Key Changes

Expand Down Expand Up @@ -189,7 +189,7 @@ This release upgrades Synapse Layer from a memory SDK to a **complete Cognitive

---

*Semantic Privacy Guard™ is now mandatory. Content is AES-256-GCM encrypted at rest.*
*Semantic Privacy Guard is now mandatory. Content is AES-256-GCM encrypted at rest.*

<!-- Tagline updated: see PUBLIC_CLAIMS_MATRIX.md -->
```
Expand Down
Loading