Paleobiology Mcp Server

Overview

Fossil biodiversity over the Paleobiology Database (PBDB), spanning roughly 540 million years. Resolve taxon fossil ranges, search fossil occurrences and collections by taxon, geologic time, and location, and plot diversity through deep time.

Tools

Resources

All resource data is also reachable via tools — the resources mirror a single-record read of paleobiology_search_occurrences / paleobiology_get_taxon / paleobiology_get_references for clients that surface resources. Tool-only clients lose nothing.

Capability reference

paleobiology_search_occurrences _tool

• Requires at least one taxon, time, geographic, formation, environment, or collection_no filter. Use base_name or base_id for a clade, taxon_name for an exact match; page with limit (default 100, maximum 10,000, subject to PBDB_MAX_OCCURRENCES) and offset. Every served row appears on both response surfaces
• Returns modern and paleo coordinates, named intervals and Ma ages, strata, and higher classification, with totalCount for PBDB's true match count. Errors: missing_filter, conflicting_taxon_filter, incomplete_bbox, inverted_ma_range, and pbdb_invalid_parameter (with PBDB's reason) for a value PBDB cannot use
• With DataCanvas on, a first page with further matches fetches and stages up to 50,000 matching occurrences from offset zero. The requested page stays inline; the notice reports staged versus total coverage. Later pages and complete pages stage nothing. Use canvas_id + table_name with paleobiology_dataframe_describe, then SQL. Reusing a canvas atomically replaces its occurrence table. At the canvas cap, reuse an existing canvas or narrow the matching set
• Optional staging shares a 55-second deadline with the page request. A transient staging-fetch failure returns the valid inline page with an unavailable notice; caller cancellation propagates

paleobiology_get_taxon _tool

• Resolve by name or taxon_no to the matched record's name, rank, classification, parent, occurrence count, and FAD/LAD range in Ma. Accepted identity (accepted_no, accepted_name, accepted_rank) and PBDB's taxonomic status are separate; use accepted_no as base_id for an accepted-clade search
• show_children adds up to 200 immediate child taxa per call; children_truncated and children_offset say whether and where to continue
• Typed errors: taxon_not_found, missing_selector, conflicting_selector when both name and taxon_no are sent

paleobiology_get_diversity _tool

• Requires a clade (base_name or base_id), optionally bounded by time, country, formation, environment, or a geographic bounding box. Choose count (genera, species, families) and resolution (period, epoch, age); defaults are genera by period
• Returns the full bin set inline, oldest-first, with sampled, implied, origination, extinction, range-through, and occurrence counts
• Originations and extinctions are first/last appearances in the selected data, not established biological events. Notices identify possible truncation at the oldest and youngest bins. recent: true extends known extant ranges to the present; the default is false
• Typed errors: missing_filter, conflicting_taxon_filter, incomplete_bbox, inverted_ma_range, pbdb_invalid_parameter for a value PBDB cannot use (an unknown interval)

paleobiology_list_intervals _tool

• Filter by name, a min_ma/max_ma overlap window, or level (eon, era, period, epoch, age); omit filters to browse the bundled ICS scale. Nonzero and single-bound windows require positive-duration overlap; an explicit equal-bound point includes both intervals meeting there
• Returns interval rank, Ma boundaries, parent_no, and an upstream scale where available. source (bundled_ics / pbdb_upstream) and snapshot_version identify the source; names outside the snapshot use PBDB
• Typed errors: interval_not_found for an unmatched name; retryable interval_lookup_unavailable when PBDB cannot resolve a name outside the snapshot; inverted_ma_range when max_ma is below min_ma

paleobiology_search_collections _tool

• Requires at least one taxon, time, geographic, country, formation, lithology, or environment filter. Page with limit (default 100, maximum 10,000, subject to PBDB_MAX_OCCURRENCES) and offset. Every served row appears on both response surfaces
• Returns locality coordinates, age, strata, lithology, environment, and n_occs, with totals. Pass collection_no to paleobiology_search_occurrences for the fauna at that locality
• Typed errors: missing_filter, conflicting_taxon_filter, incomplete_bbox, inverted_ma_range, pbdb_invalid_parameter for a value PBDB cannot use (an unknown interval, or an unknown lithology sent as the only filter)

paleobiology_get_references _tool

• Pass a nonempty list of positive integer reference_nos from occurrence or collection rows; the first 500 distinct IDs resolve in one PBDB request, and the rest come back in remaining_reference_nos for the next call
• Returns references in request order: citation, title, and publication converted from PBDB HTML to plain text with one entity decode, plus year, authors, editors, volume, issue, pages, publisher, language, and doi where recorded
• Ids PBDB has no reference for land in not_found with a notice — a miss is a result, not an error

paleobiology_dataframe_query _tool

• Runs a read-only SQL SELECT against occurrence sets staged by paleobiology_search_occurrences, using the table_name it returned; writes and file-reading functions are rejected
• The classification column is JSON — roll up by rank with json_extract_string(classification, '$.family') (also $.phylum, $.class, $.order, $.genus)
• Returns every row supplied by the canvas query. truncated: true means row_count stopped at the framework's row cap; use a stable ORDER BY with SQL LIMIT/OFFSET to continue
• Listed only when DataCanvas is on (CANVAS_PROVIDER_TYPE=duckdb)

paleobiology_dataframe_describe _tool

• Lists the tables staged on a canvas, each with its row count and column names/types/nullability — call before paleobiology_dataframe_query to discover identifiers
• Listed only when DataCanvas is on (CANVAS_PROVIDER_TYPE=duckdb)

paleobiology_dataframe_drop _tool

• Drops one staged table by canvas_id + table_name to free memory before its TTL expires; dropping a nonexistent table returns dropped: false, not an error
• Opt-in — needs DataCanvas on and PALEOBIOLOGY_DATAFRAME_DROP_ENABLED=true
• Typed error: canvas_not_found when the canvas has expired or never existed

paleobiology://occurrence/{occurrence_no} _resource

• Path param occurrence_no is a bare positive integer (regex-validated), from paleobiology_search_occurrences output
• Returns the same full occurrence detail as the tool — accepted/identified names, age, modern + paleo coordinates, formation/strata, locality — plus a CC BY 4.0 attribution field
• Typed error: occurrence_not_found

paleobiology://taxon/{taxon_no} _resource

• Path param taxon_no is a bare positive integer (regex-validated), from paleobiology_get_taxon or an occurrence's accepted_no
• Mirrors paleobiology_get_taxon's shape exactly — matched and accepted identities, status, classification, parent, FAD/LAD range — plus a CC BY 4.0 attribution field
• Typed error: taxon_not_found

paleobiology://reference/{reference_no} _resource

• Path param reference_no is a bare positive integer (regex-validated), from an occurrence or collection row
• Returns the same record as one paleobiology_get_references entry plus a CC BY 4.0 attribution field
• Typed error: reference_not_found

Features

Built on @cyanheads/mcp-ts-core: stdio and Streamable HTTP transports, pluggable auth (none / jwt / oauth), swappable storage (in-memory, filesystem, Supabase, Cloudflare KV/R2/D1), structured logging with optional OpenTelemetry tracing.

PBDB-specific:

• Type-safe client for the Paleobiology Database (PBDB) REST API. Most endpoints use readable vocab=pbdb; taxon endpoints use compact records to preserve matched rank separately from accepted rank
• Bundled ICS geologic time-scale snapshot — paleobiology_list_intervals resolves the international scale's named intervals ↔ absolute Ma boundaries with no network call, falling back to a PBDB lookup for sub-stage and regional names
• DataCanvas spill for broad occurrence queries (on by default; CANVAS_PROVIDER_TYPE=none turns it off): an inline preview plus a staged table queryable with read-only SQL (count by interval, group by formation/country, roll up by family from the classification JSON column)
• No auth, no API key — PBDB is fully open (MCP_AUTH_MODE defaults to none)
• Diversity counts reflect sampling and collection effort, not true past diversity

Agent-friendly output:

• Two coordinate systems on every occurrence — modern lng/lat and paleo lng/lat, distinctly labeled, so an agent never plots a deep-time fossil on a modern coastline
• Both temporal representations on every age — the named interval and its Ma boundaries
• Provenance on every row: reference_no for paleobiology_get_references, plus CC BY attribution on tools and resources
• Sparse upstream fields (paleo-coords, formation, late_interval) are omitted rather than zeroed

Occurrence, collection, and diversity searches accept country as comma-separated two-letter country codes or PBDB continent codes, with leading ! exclusion and ^ subtraction. Formation filters preserve PBDB's stored-name, list, and wildcard matching; do not automatically remove words such as “Formation” or “Member”.

The bundled paleobiology-guide covers identifier chaining, coverage checks, sampled turnover, and citation workflows. It ships with plugins, npm, Docker, and MCPB, and is served through MCP skill resources.

Getting started

Connect to the hosted instance:

claude mcp add --transport http paleobiology-mcp-server https://paleobiology.caseyjhand.com/mcp

Or run it locally:

claude mcp add paleobiology-mcp-server -- npx -y @cyanheads/paleobiology-mcp-server

For other clients, add this to the MCP config file:

{  "mcpServers": {    "paleobiology-mcp-server": {      "command": "npx",      "args": ["-y", "@cyanheads/paleobiology-mcp-server"]    }  }}

PBDB is keyless. For Docker or a source build, see Running the server.

SQL over large occurrence sets is on by default (the @duckdb/node-api peer dep ships in dependencies). With CANVAS_PROVIDER_TYPE=none, paleobiology_search_occurrences still returns its complete requested page, and the paleobiology_dataframe_* tools are left off tools/list (the HTTP landing page shows them as disabled, with the setting that enables them).

Configuration

All variables are optional — the server runs with no configuration against the public PBDB API.

See .env.example for the full list of optional overrides.

Running the server

Local development

Requires Bun v1.4 or higher (or Node.js v24+). No API key is required.

git clone https://github.com/cyanheads/paleobiology-mcp-server.gitcd paleobiology-mcp-serverbun installcp .env.example .env # optional overrides

• Build and run:

  sh

  # One-time buildbun run rebuild
  # Run the built serverbun run start:stdio# orbun run start:http

• Run checks and tests:

  sh

  bun run devcheck   # Lint, format, typecheck, securitybun run test       # Vitest test suitebun run lint:mcp   # Validate MCP definitions against spec

Docker

docker run --rm -p 3010:3010 ghcr.io/cyanheads/paleobiology-mcp-server:latest

The image defaults to HTTP transport, stateless session mode, and logs to /var/log/paleobiology-mcp-server. For stdio in an MCP client, use "command": "docker" and "args": ["run", "-i", "--rm", "-e", "MCP_TRANSPORT_TYPE=stdio", "ghcr.io/cyanheads/paleobiology-mcp-server:latest"]. Build locally with docker build -t paleobiology-mcp-server .; OpenTelemetry peers are installed unless built with --build-arg OTEL_ENABLED=false.

Project structure

Development guide

See CLAUDE.md/AGENTS.md for development guidelines and architectural rules. The short version:

• Handlers throw, framework catches — no try/catch in tool logic
• Use ctx.log for request-scoped logging, ctx.state for tenant-scoped storage
• Register new tools and resources via the barrels in src/mcp-server/*/definitions/index.ts
• Wrap external API calls: validate raw → normalize to domain type → return output schema; never fabricate missing fields (a missing paleo-coordinate is "unknown", not 0,0)

Data attribution

Data is from the Paleobiology Database, licensed CC BY 4.0 — credit it in downstream use.

Contributing

Issues are welcome. Run checks and tests before submitting:

bun run devcheckbun run test

License

Apache-2.0 — see LICENSE for details.