<aside> 🧠
Single-path output. Integrates the TAO Optimization Inversion framework (TAO Paper Revision — OSDG Experiment Draft (v2.2)), Yennefer thermodynamic routing (‣), architect-tier invariants (Greg × /maru — Architect Review), and the UCP Decision Matrix (Genesis Conductor × UCP — Integration Spec & Model Attainment Decision Matrix) into a unified agent-based intent router backed by Dissonance Memory MCP for cross-session state and TAO-derived energy-aware scheduling.
</aside>
<aside> 🔗
Canonical cluster anchors (previewable). Yennefer cluster · TAEX · Retraining loop · §6.17 · Integrations registry · Integrative Flow.
‣ · ‣ · ‣ · ‣ · ‣ · ‣ · ‣ · ‣
</aside>
<aside> 🛰️
A2A Operationalization · Surface 2.2 (Intent router). Canonical index: A2A Operationalization Index — Spine Footers, JSONL Prompts, Model Variants, Permissions.
mcp.genesisconductor.io/memory/dissonance Worker + D1 + KV; wire dissonance_recall + thermodynamic_route UCP extensions.{"v":"a2a/1.0","kind":"intent.route","id":"taex.route.001","surface":"<https://www.notion.so/c33eb0932f0844a8a2a9cd631e180e30","input":{"intent_text":"><raw_intent>","workspace_id":"<ws>","policy_tier":"T2"},"pipeline":["classify","memory.recall","memory.dissonance_check","decision_matrix","yennefer.eta_thermo","aal.submitTask","memory.store"],"model":{"classifier":"claude-3-5-sonnet","guardian":"claude-opus-4-1","reframe":"claude-3-5-sonnet"},"exec":"cloudflare_worker:gc-mcp-server","capsule_event":"taex_routing.routed"}
</aside>
TAEX = TAO + Attestor + EXtensions. It is the composition of:
no_win_reframe, gas_bound_quote, prompt_eval, prompt_promote, thermodynamic_route, dissonance_recall) that fire conditionally based on intent classification.TAEX is the runtime identity of the Genesis Conductor mesh when operating in intent-routed mode.
Dissonance Memory is a Model Context Protocol (MCP) server that provides cross-session, cross-agent episodic and semantic memory with built-in cognitive dissonance detection. When an agent's current intent conflicts with prior commitments, beliefs, or workspace state stored in memory, the Dissonance layer surfaces the conflict as a structured signal before routing proceeds.
MCP surface: mcp.genesisconductor.io/memory/dissonance
Core operations:
| MCP Tool | Description | Returns |
|---|---|---|
memory.recall |
Retrieve episodic + semantic memories relevant to the current intent | { memories[], relevance_scores[], dissonance_flags[] } |
memory.store |
Persist a new memory entry (event, decision, commitment, belief) | { memory_id, stored_at } |
memory.dissonance_check |
Compare a proposed action against stored commitments and beliefs | { dissonance_score, conflicts[], resolution_suggestions[] } |
memory.project |
Build a state projection from memory for a given workspace/scope | { projection, confidence, staleness_ms } |
memory.forget |
Soft-delete a memory (mark as superseded, retain audit trail) | { forgotten_id, reason } |
Dissonance scoring:
$D(a, M) = \max_{m \in M} \text{conflict}(a, m) \times \text{recency}(m) \times \text{confidence}(m)$
where $a$ is the proposed action, $M$ is the memory set, and $D > 0.3$ triggers a Guardian review before routing proceeds.