MCP server for pydna - Python DNA simulation library. Provides comprehensive DNA simulation and molecular cloning tools through the Model Context Protocol, including PCR simulation, primer design, restriction analysis, DNA assembly, and gel electrophoresis simulation.
14 tools with consistent naming (verb_noun pattern: create_, get_, design_, etc.) and descriptions present. However, descriptions are generic and lack LLM-optimization guidance (e.g., 'Get detailed information' vs 'Retrieve GC content, melting temperature, and molecular weight'). Input schemas visible with types and descriptions, but output schemas are inferred from Pydantic models rather than explicitly documented in tool registration. No error handling guidance, no parameter constraints (enums, ranges), and no indication of which tools are idempotent or safe to retry. Parameter descriptions lack format/range details (e.g., 'target_tm' has no units or valid range). One tool (pydna_write_sequence_file) is destructive but lacks confirmation or dry-run pattern.
Analyze primer annealing to template
Assemble DNA fragments by homologous recombination
Create a DNA sequence from string and return comprehensive sequence information.
Design PCR primers for a DNA sequence
Digest DNA sequence with restriction enzymes
Download sequence from GenBank
Simulate gel electrophoresis
Design fragments for Gibson assembly
Descriptions lack LLM-optimization guidance. Generic phrases like 'Get detailed information' and 'Analyze restriction enzyme cutting sites' do not explain WHEN to use the tool, what it returns, or how it differs from similar tools. Baseline: A+ tools have 50-200 char descriptions answering what/when/why.
Parameter descriptions lack format/range constraints. 'target_tm' has no units (Celsius assumed but not stated), no valid range (e.g., 50-80°C), and no guidance on typical values. 'overlap' in gibson_assembly has no minimum/maximum. Baseline: A+ tools specify min/max for numeric params and format for strings.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 45 | <=2025-11-25 | v2 |
Simulate PCR amplification with given primers and template
Read DNA sequence from file (FASTA, GenBank, etc.)
Analyze restriction enzyme cutting sites
Get reverse complement of a DNA sequence
Get detailed information about a DNA sequence
Write DNA sequence to file
Output schemas are inferred from Pydantic models (SequenceInfo, PrimerInfo, etc.) but not explicitly documented in tool registration. LLMs cannot see what fields to expect without reading implementation code. Baseline: 100% of A+ tools have documented return types.
Destructive tool (pydna_write_sequence_file) lacks confirmation or dry-run pattern. No error handling guidance for file I/O failures (permission denied, disk full, invalid path). Agents cannot know if a write succeeded or how to recover.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). Agents cannot distinguish safe tools from destructive ones without reading descriptions. pydna_write_sequence_file should be marked destructive; most others should be marked readOnly.