| name | reactome-search |
| description | Query Reactome Content and Analysis Services: pathway enrichment of a gene list, identifier mapping, reaction participants, pathway hierarchy, and knowledgebase search. Use for pathway analysis and enrichment. Public API, no credential needed. |
| license | Apache-2.0 |
Reactome Search (Codex-native)
Codex-native adaptation of Google DeepMind's reactome-database skill
(science-skills, Apache-2.0).
Uses the public API directly through Codex's own tools.
Gates (ask before proceeding)
- Terms notice: on first use in a workspace, tell the user to review the
source's terms of use and record the notice in provenance.
- Network / install: read-only public access; be polite about rate limits.
Write large responses to a file and parse them; if you write helper code, run
it via
uv run, never bare python3. Ask before installs or large downloads.
Endpoints
- Reactome ContentService
https://reactome.org/ContentService
- Reactome AnalysisService
https://reactome.org/AnalysisService
Workflow
Search pathways and reactions first with science_search_reactome; use the
services directly for details, participants, and enrichment.
- For enrichment, POST the identifier list to
/AnalysisService/identifiers/ and read results by token.
- For content, query pathways/reactions/participants via the ContentService endpoints.
- Provenance & review: record queries, endpoints, and cited identifiers with
$science-provenance; check claims with $science-review before presenting.
Boundaries
- Report the species and the analysis token; enrichment p-values depend on the background.
- List not-found/unmapped identifiers honestly; cite only stable ids actually returned.
