EcoSeek
EcoSeek logo

EcoSeek

Scientific Agent Environment for Ecology

EcoSeek is an open-source scientific agent environment for ecological and computational biology workflows. It combines domain-specialized language models, ecological tool pipelines, and a secure gateway to help researchers run species distribution models, analyze biodiversity data, and build knowledge graphs from their laptop or an HPC cluster.

View on GitHub EcoAgent Tools Try our Demo ✨

Three ways to use EcoSeek

FREE

DIY / Community

Run everything locally with EcoCoder and EcoAgent. Free, reproducible, and fully open. Uses local Ollama inference and 30+ ecological tools. No API keys needed.

BYOK

Bring Your Own Key

Use your own API key for a compliant cloud provider (Arcee, OpenAI, etc.) when you lack local GPU hardware. Keys are stored locally and never leave your machine. Combine with EcoAgent tools for hybrid workflows.

LAB

Lab / Managed

Connect to lab servers, workstations, and HPC clusters through AgenticPlug. Audit trails, approval gates for compute operations, and persistent sessions. Designed for multi-user research environments.

Architecture

EcoSeek client (UI / CLI / agent runtime) | v AgenticPlug gateway (auth, sessions, policy, approvals, audit) | v Compute backends EcoCoder local ---- domain-specialized ecological LLM EcoCoder cluster -- remote HPC inference via tunnel EcoAgent tools ---- 25 ecological analysis tools Emily / Hermes ---- DiDAL v2 dialectical protocol Alpha (local) ↔ Beta (remote) BYOK keystore ---- user-provided API key (non-Chinese providers) HPC / Hermes ------ lab servers and clusters
GitHub proves who the user is. AgenticPlug decides what the user can do.

Components

  • AgenticSeek User-facing client foundation. Provides the agent runtime, CLI, and web UI. EcoSeek adds ecological providers (EcoCoder local/cluster) and connector discovery. BYOK accepts non-Chinese providers only.
  • AgenticPlug Secure gateway and connector layer. Handles authentication (GitHub identity), scoped sessions, role-based authorization, approval workflows for risky operations, and audit logging.
  • EcoCoder Domain-adapted ecological code LLM (fine-tuned on ~80K lines of ecological computing code; retrain on a compliant base in progress — the Qwen-based 7B release is deprecated under the no-Chinese-AI policy). Exposes an OpenAI-compatible inference endpoint.
  • EcoAgent Ecological workflow and tool layer. 25 tools including species queries (GBIF), literature search, diversity analysis, species distribution modeling (MaxEnt, SDM), host-parasite interaction extraction, taxonomy resolution, and knowledge graph construction.
  • Emily Scientific reasoning engine with DiDAL v2 dialectical protocol. Alpha (local, Emily) acts as naive scientific interlocutor; Beta (remote, Hermes) is the expert researcher. Automatic complexity-based routing (direct → didal → didal_literature), structured evidence retrieval, critique-revise loops, and mini-report generation.
  • Hermes Remote orchestration service at hermes.ecoseek.org. Optional advanced scientific reasoning backend. Runs on reumanlab with Trinity Large Thinking (Arcee) and EcoAgent tools. Emily escalates complex tasks to Hermes for expert synthesis.

Ecological Tools

EcoAgent exposes 25 tools via an HTTP API with risk classification and AgenticPlug approval gating:

  • query_species
  • query_papers
  • compute_diversity
  • fit_sdm
  • fit_maxent
  • extract_triplets
  • resolve_taxonomy
  • build_knowledge_graph
  • compute_bioclim
  • classify_abstract
  • query_gbif_parquet
  • evaluate_niche
  • predict_susceptibility
  • compute_effort_bias
  • query_pubtator
  • query_cofid
  • compute_ecological_distance
  • fit_geotax
  • search_literature

Read-only tools run freely. Compute-intensive tools (SDM, MaxEnt, niche evaluation) carry risk metadata and can require approval through AgenticPlug before execution.

Write literature reviews with EcoSeek

EcoSeek is the factory side of your literature review. Deploy a scientific agent that searches the literature, gathers evidence, and drafts — then you, the scientist, stay the author. The result is a citable preprint, not an AI dump.

  1. Search. The EcoAgent literature tools (search_literature, query_papers, classify_abstract) pull and rank relevant papers across your topic.
  2. Synthesize. Emily's DiDAL protocol runs critique-revise loops, escalating complex synthesis to Hermes for expert reasoning — you review and edit every step.
  3. Publish. Submit your review to litreview.org, where it becomes a citable preprint with versioning, citation metadata, and an openly declared AI-assistance record — the scientist is the author, AI is the assistant, without apology.
litreview.org — our preprint platform for expert-written, AI-assisted literature reviews. Every published review is released under CC BY 4.0 for barrier-free citation and reuse.

Getting Started

DIY / Community mode

Run the EcoAgent tool server locally:

pip install ecoagent[full]
python -m ecoagent.tool_server --port 8200

# Or with Docker
docker run -p 8200:8200 -e ECOAGENT_PROFILE=ci ghcr.io/alrobles/ecoagent:0.1.0

BYOK mode

Store your API key locally (non-Chinese providers only), then configure EcoSeek:

python -m sources.keystore set arcee YOUR_KEY
# Key stored in OS keychain or ~/.config/ecoseek/keys.json
# Never committed, logged, or sent to GitHub

Lab / Managed mode

Deploy AgenticPlug as the gateway, connect to your HPC cluster:

cd agenticplug && npm start
# Configure connectors, sessions, and approval policies
# See docs/approval-workflow.md for risky operation gates
Attribution: EcoSeek is built on a fork of AgenticSeek and uses Hermes Agent for orchestration. We gratefully acknowledge both projects and their contributors as foundations for this work. EcoSeek is an independent downstream adaptation focused on scientific and ecological computing. EcoSeek is licensed under the GNU AGPLv3. AgenticSeek is licensed under GPLv3; Hermes Agent is licensed under MIT.

Contact

github.com/alrobles · [email protected]

EcoSeek is developed at the University of Kansas.