Elnora Agent Capabilities
The Elnora Agent is a sandboxed Python environment with ~78 core tools + 2,100 ToolUniverse scientific tools. Interact via tasks create and tasks send — describe what you need in plain language.
Tool Access
This skill documents what the Elnora Agent can do. It does not have dedicated CLI commands — interact with the agent through the elnora-tasks skill.
Elnora is available via two methods. Use whichever is configured.
Option A — CLI via Bash (preferred)
Run commands via your Bash/Shell tool as elnora tasks create ... or elnora tasks send .... Verify with elnora --version.
Tip: Add --stream for real-time agent output:
elnora --compact tasks send <TASK_ID> --message "..." --stream
Option B — MCP tools (when CLI isn't installed)
Look for tools prefixed mcp__elnora__ in your available tools — specifically mcp__elnora__elnora_tasks_create and mcp__elnora__elnora_tasks_send. Call them with structured parameters (camelCase).
If neither is available, tell the user to install one:
- CLI:
curl -fsSL https://cli.elnora.ai/install.sh | bash (macOS/Linux)
or irm https://cli.elnora.ai/install.ps1 | iex (Windows)
- MCP:
claude mcp add elnora --transport http --scope user https://mcp.elnora.ai/mcp
then /mcp to authenticate.
Never fabricate tool names like elnora_run_agent or elnora_ask_agent. The agent is interacted with via elnora_tasks_create, elnora_tasks_send, and the aggregate elnora_protocols_generate — see the elnora-tasks skill for full command reference.
Invocation
CLI="elnora"
Quick Start
# Create a task and wait for the response
$CLI --compact tasks create --project <PROJECT_ID> --title "My task" --message "Your request"
# Send follow-up and wait for response
$CLI --compact tasks send <TASK_ID> --message "Follow-up request" --wait
# Or stream the response in real-time
$CLI --compact tasks send <TASK_ID> --message "Follow-up request" --stream
What the Agent Can Do
| Capability |
Examples |
| Web search (34 tools) |
Real-time search, neural/semantic search, deep research, URL extraction, site crawling. Providers: Tavily, Exa, Valyu, Perplexity |
| Academic databases (12 tools) |
PubMed, ArXiv, Semantic Scholar, bioRxiv, Europe PMC, OpenAlex, UniProt, ClinicalTrials.gov, ChEMBL, Wolfram Alpha |
| 2,100+ scientific tools (ToolUniverse) |
Protein structure (AlphaFold, PDB), genomics (Ensembl, ClinVar), chemistry (PubChem, DrugBank), pathways (KEGG, Reactome), drug safety (OpenFDA), and 21 more categories |
| 35 domain skills |
Literature review, experimental design, drug discovery workflow, protein engineering, single-cell RNA QC, statistical analysis, scientific writing |
| File operations (11 tools) |
Create/read/search files, full-text grep, upload attachments, link files to tasks |
| Memory (9 tools) |
Remember facts across tasks, share findings between agents, recall prior context |
| Code execution |
Persistent Python REPL with pandas, numpy, biopython. Variables survive across executions. 30s timeout, 1MB output max |
Example Prompts
# Web research
$CLI --compact tasks send "$TASK" --message "Search for recent CRISPR delivery methods and summarize" --wait
# Literature review
$CLI --compact tasks send "$TASK" --message "Search PubMed for BRCA1 DNA repair papers from 2024" --wait
# Drug target research
$CLI --compact tasks send "$TASK" --message "Search for compounds targeting EGFR, cross-reference with active clinical trials" --wait
# Scientific computation
$CLI --compact tasks send "$TASK" --message "Use ToolUniverse to run AlphaFold on this sequence: MVLSPADKTNVKAAWGKVGA" --stream
# Memory
$CLI --compact tasks send "$TASK" --message "Remember that our lab uses Q5 polymerase for all high-fidelity PCR at 62C" --wait
# Reference existing files
$CLI --compact tasks send "$TASK" --message "Read the attached template and generate a new version" --file-refs "<FILE_ID>" --wait
1---2name: elnora-agent3description: This skill should be used when the user asks about "Elnora agent capabilities", "what can the agent do", "what tools does Elnora have", "what can Elnora do", "agent tools", "web search", "academic search", "PubMed", "ArXiv", "Exa", "Tavily", "Perplexity", "Valyu", "ToolUniverse", "scientific tools", "agent memory", "code execution", "sandbox", "search papers", "search literature", "drug discovery", "protein analysis", "clinical trials", "file operations", "agent skills", or any question about what the Elnora AI Agent can do when you send it a task.4---56# Elnora Agent Capabilities78The Elnora Agent is a sandboxed Python environment with ~78 core tools + 2,100 ToolUniverse scientific tools. Interact via `tasks create` and `tasks send` — describe what you need in plain language.910## Tool Access1112This skill documents **what the Elnora Agent can do**. It does not have dedicated CLI commands — interact with the agent through the `elnora-tasks` skill.1314Elnora is available via two methods. Use whichever is configured.1516**Option A — CLI via Bash (preferred)**1718Run commands via your Bash/Shell tool as `elnora tasks create ...` or `elnora tasks send ...`. Verify with `elnora --version`.1920**Tip:** Add `--stream` for real-time agent output:2122 elnora --compact tasks send <TASK_ID> --message "..." --stream2324**Option B — MCP tools (when CLI isn't installed)**2526Look for tools prefixed `mcp__elnora__` in your available tools — specifically `mcp__elnora__elnora_tasks_create` and `mcp__elnora__elnora_tasks_send`. Call them with structured parameters (camelCase).2728**If neither is available, tell the user to install one:**2930- CLI: `curl -fsSL https://cli.elnora.ai/install.sh | bash` (macOS/Linux)31 or `irm https://cli.elnora.ai/install.ps1 | iex` (Windows)32- MCP: `claude mcp add elnora --transport http --scope user https://mcp.elnora.ai/mcp`33 then `/mcp` to authenticate.3435**Never fabricate tool names** like `elnora_run_agent` or `elnora_ask_agent`. The agent is interacted with via `elnora_tasks_create`, `elnora_tasks_send`, and the aggregate `elnora_protocols_generate` — see the `elnora-tasks` skill for full command reference.3637## Invocation3839```bash40CLI="elnora"41```4243## Quick Start4445```bash46# Create a task and wait for the response47$CLI --compact tasks create --project <PROJECT_ID> --title "My task" --message "Your request"4849# Send follow-up and wait for response50$CLI --compact tasks send <TASK_ID> --message "Follow-up request" --wait5152# Or stream the response in real-time53$CLI --compact tasks send <TASK_ID> --message "Follow-up request" --stream54```5556## What the Agent Can Do5758| Capability | Examples |59|------------|----------|60| **Web search** (34 tools) | Real-time search, neural/semantic search, deep research, URL extraction, site crawling. Providers: Tavily, Exa, Valyu, Perplexity |61| **Academic databases** (12 tools) | PubMed, ArXiv, Semantic Scholar, bioRxiv, Europe PMC, OpenAlex, UniProt, ClinicalTrials.gov, ChEMBL, Wolfram Alpha |62| **2,100+ scientific tools** (ToolUniverse) | Protein structure (AlphaFold, PDB), genomics (Ensembl, ClinVar), chemistry (PubChem, DrugBank), pathways (KEGG, Reactome), drug safety (OpenFDA), and 21 more categories |63| **35 domain skills** | Literature review, experimental design, drug discovery workflow, protein engineering, single-cell RNA QC, statistical analysis, scientific writing |64| **File operations** (11 tools) | Create/read/search files, full-text grep, upload attachments, link files to tasks |65| **Memory** (9 tools) | Remember facts across tasks, share findings between agents, recall prior context |66| **Code execution** | Persistent Python REPL with pandas, numpy, biopython. Variables survive across executions. 30s timeout, 1MB output max |6768## Example Prompts6970```bash71# Web research72$CLI --compact tasks send "$TASK" --message "Search for recent CRISPR delivery methods and summarize" --wait7374# Literature review75$CLI --compact tasks send "$TASK" --message "Search PubMed for BRCA1 DNA repair papers from 2024" --wait7677# Drug target research78$CLI --compact tasks send "$TASK" --message "Search for compounds targeting EGFR, cross-reference with active clinical trials" --wait7980# Scientific computation81$CLI --compact tasks send "$TASK" --message "Use ToolUniverse to run AlphaFold on this sequence: MVLSPADKTNVKAAWGKVGA" --stream8283# Memory84$CLI --compact tasks send "$TASK" --message "Remember that our lab uses Q5 polymerase for all high-fidelity PCR at 62C" --wait8586# Reference existing files87$CLI --compact tasks send "$TASK" --message "Read the attached template and generate a new version" --file-refs "<FILE_ID>" --wait88```
Elnora-AI