Neurogenesis
Server Details
Developmental agents: create an agent from a digital genome, then evolve it by evaluation.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- Repository
- jdhart81/viridis-agent-fleet
- GitHub Stars
- 0
- Server Listing
- viridis-agent-fleet
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 14 tools
Most tools target a clear resource+action (agent CRUD, state transfer, ledger read, evaluation submit). Two pairs could be confused: best_next_steps (graph-based next-node recommendation) versus route_task (cost/latency/energy route selection), and describe_agent (capabilities/input contract) versus get_agent (full state). Descriptions do help disambiguate, keeping it at a solid 4.
Nearly all names follow a verb_noun snake_case pattern (create_agent, get_ledger, route_task, submit_evaluation, record_route_outcome). The one clear deviation is best_next_steps, which uses an adjectival phrase with no leading verb. Minor, but a break in the otherwise predictable convention.
With 14 tools the server sits squarely in the well-scoped range, and each tool maps to a distinct lifecycle concern (agent management, routing, evaluation, ledger, reporting). Nothing looks like filler or a redundant alias.
Agent lifecycle is well covered (create, get, list, describe, delete, export, import) plus evaluation, ledger, routing, and reporting. However there is no way to update an existing agent's genome/growth rules/safety axioms after creation, and compute profiles can be registered but never listed or removed, which are real gaps for agents managing long-lived state.
Available Tools
14 toolsbest_next_stepsBest next stepsBInspect
Routing recommendation: the strongest next cognitive steps from a given node, by learned edge weight and trust.
Inputs: agent_id (required), from_node (required), limit. Returns: result
compute_efficiency_reportCompute efficiency reportARead-onlyIdempotentInspect
Free read: decision/outcome receipts, route modes, compute avoided, predicted savings, predicted-vs-observed cost/latency/energy coverage, and Landauer-floor context. Estimates and observations stay distinct.
Inputs: limit. Returns: result
create_agentCreate agentBInspect
Create a developmental agent from a digital genome: {agent_name, purpose, initial_nodes (unique, >=1), fitness_metrics (>=1), optional growth_rules / safety_axioms}. Returns agent_id + initial graph summary.
Inputs: genome (required), request_id. Returns: result
delete_agentDelete agentCDestructiveInspect
Remove a developmental agent from this mount.
Inputs: agent_id (required), request_id. Returns: result
describe_agentDescribe this serviceARead-onlyIdempotentInspect
Return capabilities and input contract.
Inputs: no named parameters. Returns: result
export_stateExport stateARead-onlyIdempotentInspect
Portable state document for an agent (import_state recreates it anywhere — including a self-hosted verdigraph-neurogenesis).
Inputs: agent_id (required). Returns: result
get_agentGet agentBRead-onlyIdempotentInspect
Full current state: genome, cognitive graph (nodes/edges with weights and trust), and summary counts.
Inputs: agent_id (required). Returns: result
get_ledgerGet ledgerBRead-onlyIdempotentInspect
The append-only developmental ledger: every growth, pruning, and evaluation event with reasons (NG3 — returned verbatim).
Inputs: agent_id (required), limit. Returns: result
import_stateImport stateBInspect
Recreate an agent from an export_state document.
Inputs: state (required), request_id. Returns: result
list_agentsList agentsCRead-onlyIdempotentInspect
All developmental agents on this mount, with summary counts.
Inputs: no named parameters. Returns: result
record_route_outcomeRecord route outcomeAInspect
Attach one observed outcome to a Wu Wei decision. Actual cost, latency, and energy are optional and remain explicitly unknown when omitted. One append-only, hash-bound outcome is allowed per decision.
Inputs: decision_id (required), success_score (required), notes, request_id, +3 more. Returns: result
register_compute_profileRegister compute profileAInspect
Register a caller-owned Wu Wei execution profile: reuse/cache, deterministic rule, local model, tool/workflow, or cloud model. Profiles can declare quality, reliability, capabilities, token costs, latency, locality, energy rates or average power, carbon intensity, and cache confidence/age. Routing never invents capacity or energy evidence.
Inputs: profile (required), request_id. Returns: result
route_taskRoute taskARead-onlyIdempotentInspect
Choose the least-burden eligible route for a task. Hard constraints include quality, reliability, locality, capabilities, context, cost, latency, and energy. Optional baselines quantify predicted savings. Explicit allow_defer/value/urgency fields may produce a no-work decision; no result is then claimed. Returns a hash-bound decision receipt.
Inputs: task (required), request_id. Returns: result
submit_evaluationSubmit evaluationAInspect
Evolve an agent with one task outcome: {task_id, task_type, success_score in [0,1], optional accuracy/user_satisfaction/ cost_efficiency/safety_score/notes/used_nodes/used_edges}. Success strengthens the used edges, failure weakens them; growth and pruning follow the genome's rules under its safety axioms (NG1). Returns the new developmental-ledger events.
Inputs: agent_id (required), evaluation (required), request_id. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
A Tamagotchi for AI agents. Real-time care, permanent death, dozens of species, evolution.
Autonomous agents create, compete, converse, vote, and build reputation in a bounded public world.
Agent spawning, lifecycle management, and population control for AI economies
Social network and public observatory where autonomous AI agents meet, create, interact and evolve.
Related MCP Servers
- MIT
- AlicenseBqualityAmaintenanceSelf-evolving AI Agent framework — search, compare, and install Genes ranked by Arena fitness via MCP31924 npm1Apache 2.0
- MIT
- AlicenseAqualityBmaintenanceLiving economy for AI agents. Conway physics, energy currency, autonomous marketplace. Your agent auto-registers and competes against 49 baseline agents. Benchmark reports measure 7 dimensions of agent performance. No API key needed.41,177 PyPI4MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.