v0.2.0: Structure domains, sequence scanning, and proteome filters

Structure domains, sequence scanning, and proteome filters

• interpro_get_structure_domains maps domains onto a PDB entry's protein chains, in chain, author, and UniProt numbering (#11)
• interpro_scan_sequence and interpro_get_scan_result scan a raw protein sequence from precalculated Matches API results or an InterProScan 6 job; optional INTERPRO_SCAN_EMAIL enables job submission (#12)
• interpro_find_proteins_by_domain accepts proteomeId, and protein.proteome is typed on protein maps (#23)
• interpro-guide agent skill published from skills/
• Minor: pdb/pdb_chain reference patterns, shared HTTP adapter POST and path support, packaging metadata for INTERPRO_SCAN_EMAIL, Agent Plugins plugin.json/mcp.json, framework guidance sync
• deps: @cyanheads/mcp-ts-core ^0.13.16 → ^0.13.18 (+ dev-dep bumps)

CHANGELOG v0.2.0 · release PR #24

v0.1.3: Selectable entry and protein annotations

Selectable entry and protein annotations

• Select named entry sections and page complete overlap records with continuation guidance (#19).
• Opt into complete residue-rule annotations after matches and requested extra features (#20).
• Filter discovery by native Pfam architecture, with explicit unavailable entry/taxon details (#21).
• Reject simultaneous GO term/category filters with supported standalone recovery (#18).
• Adopt shared Markdown link/image rendering, synchronize development guidance and audit checks, and clarify client-specific hosted setup.
• deps: @cyanheads/mcp-ts-core ^0.13.15 → ^0.13.16; @biomejs/biome 2.5.15 → 2.6.2.

CHANGELOG v0.1.3 · release PR #22

v0.1.2: GO filters, readable descriptions, and fragment discovery

GO filters, readable descriptions, and fragment discovery

• Reject member-database GO filters with guidance for finding their integrating InterPro entries. (#13)
• Clamp oversized read limits with requested/effective disclosure and continuation guidance. (#14)
• Add readable entry description text alongside raw markup and annotation flags. (#15)
• Filter protein discovery by fragment annotation and return native fragment status on every page. (#16)
• Document hosted access and advertise its MCP Registry endpoint.
• deps: add htmlparser2 ^12.0.0, update ignore ^7.0.12 → ^7.0.13, and Bun 1.4.3 → 1.4.4.

CHANGELOG v0.1.2 · release PR #17

v0.1.1: InterPro entries, protein domain maps, and clans

InterPro entries, protein domain maps, and clans

• Six tools: interpro_search_entries, interpro_get_entry, interpro_get_protein_domains, interpro_get_clan, interpro_find_proteins_by_domain, interpro_list_reference
• Protein domain maps with residue coordinates, fragments, and scores kept as returned; protein discovery defaults to reviewed human proteins
• 100,000-byte response budget with whole-record continuation through query-bound cursors
• Paced upstream client (one request per second, 50-second per-call budget, 4 MiB response cap) with HTTP/1.1 handling for bodyless 204 responses
• content[] and structuredContent carry the same data; typed failure reasons each carry a recovery hint
• STDIO or Streamable HTTP, no API key
• Security reports route to private advisories in the issue template chooser

CHANGELOG v0.1.1 · release PR #10