encode-ccres-search

name	encode-ccres-search
description	Query the ENCODE Registry of candidate cis-Regulatory Elements (cCREs) via the SCREEN GraphQL API, and ENCODE experiments/files via the ENCODE Portal REST API, across human cell types. Public APIs, no credential needed.
license	Apache-2.0

Encode Ccres Search (Codex-native)

Codex-native adaptation of Google DeepMind's encode-ccres-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

• SCREEN GraphQL https://factorbook.api.wenglab.org/graphql
• ENCODE Portal REST https://www.encodeproject.org/ (append ?format=json)

Workflow

1. For regulatory annotations, query cCREs by region/cell type via the SCREEN GraphQL endpoint.
2. For raw experimental data (ChIP-seq peaks, etc.), query the ENCODE Portal REST search with ?format=json and restrict fields.

• Provenance & review: record queries, endpoints, and cited identifiers with $science-provenance; check claims with $science-review before presenting.

Boundaries

• Report the genome assembly and biosample/cell type; regulatory calls are cell-type specific.
• Cite only ENCODE accessions actually returned.