Add Ensembl Bacteria tools for bacterial genome queries

[citizenlee]

Summary

• Adds 3 new bacteria-specific MCP tools (gget_bacteria_species, gget_bacteria_search, gget_bacteria_ref) that query the Ensembl REST API directly, since the gget library explicitly excludes bacteria from search() and ref()
• New EnsemblBacteriaClient class in ensembl_bacteria.py handles species discovery, gene symbol resolution via xrefs, and FTP URL construction for the collection-based bacterial genome layout
• Existing gget_info and gget_seq tools already work with bacterial Ensembl IDs — these new tools provide the missing discovery layer

New Tools

Tool	Purpose	API Used
gget_bacteria_species	Find full Ensembl species name from partial input	GET /info/genomes/taxonomy/{taxon} + species index fallback
gget_bacteria_search	Search bacterial genes by symbol	GET /xrefs/symbol/ + POST /lookup/id
gget_bacteria_ref	Get FTP download URLs for bacterial reference genomes	Species index TSV parsing + FTP directory scraping

Workflow Example

bacteria_species("escherichia coli k-12") → full species name
bacteria_search(["dnaA"], species=full_name) → gene ID "b3702"
gget_info("b3702") → gene metadata (existing tool, works as-is)
gget_seq("b3702") → nucleotide sequence (existing tool, works as-is)
bacteria_ref(species=full_name, which="dna") → FTP URLs for genome FASTA

Test plan

• 19 integration tests covering species search, gene search, reference URLs, error cases, MCP server wrappers, and end-to-end workflow (all passing against live Ensembl API)
• Existing test suite passes with no regressions (5/5 in test_gget_simple.py)
• Verify tools appear correctly when mounted in Cultivarium MCP

🤖 Generated with Claude Code