Name: ontseq-platform
Author: xhuljanoshehu

Imported from xhuljanoshehu/ontseq-platform (AGENTS.md). Install upstream with npx skills add xhuljanoshehu/ontseq-platform. Copyright stays with the author.

Repository guardrails for coding agents

This project processes potentially identifying human genomic data. Automated contributors must preserve the following boundaries:

Never add patient data, genomic run files, credentials, direct identifiers or reports to Git. Use synthetic fixtures only.
Treat every ISCN string as an expert-reviewable proposal. Do not remove the research-use warning or introduce automatic clinical release.
Keep assay-specific QC and reportability rules versioned. A caller default is not a validated clinical threshold.
Any change that can alter biological output needs tests, provenance fields, a changelog entry and validation-impact review.
Prefer typed adapters and structured contracts over parsing presentation files.
Run make safety, make versions, make lint, and make test before proposing a change.
Do not add a public license, public deployment, cloud upload or external data transfer without explicit owner and institutional approval.
Before acquiring external data files, check the exact source, access/use conditions and technical risks using primary evidence. Distinguish metadata checks from inspection of actual file contents, state unresolved risks, and never equate public availability with anonymity, malware-free contents or biological validity. This check does not introduce an additional approval requirement beyond the existing project rules.

The source of truth for scope and limitations is docs/ARCHITECTURE.md, docs/EVIDENCE_BASE.md, and docs/CLINICAL_VALIDATION.md. The thesis traceability file is historical context only and must not be treated as a technical specification.

Install ontseq-platform and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "xhuljanoshehu-ontseq-platform-ontseq-platform", "kind": "skill", "name": "ontseq-platform", "description": "This project processes potentially identifying human genomic data. Automated contributors must preserve the following boundaries:", "publisher": "xhuljanoshehu", "version": "1.0.0", "capabilities": { "domains": [ "general" ], "tags": [ "skill-md", "github-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "This project processes potentially identifying human genomic data. Automated contributors must preserve the following boundaries:" ], "primary": false, "metadata": { "source": { "provider": "github-agents", "repository": "https://github.com/xhuljanoshehu/ontseq-platform", "path": "AGENTS.md", "ref": "f2776f13b16d31980774a070663bc2a0b42e90ad", "url": "https://github.com/xhuljanoshehu/ontseq-platform/blob/f2776f13b16d31980774a070663bc2a0b42e90ad/AGENTS.md", "key": "xhuljanoshehu/ontseq-platform/AGENTS.md" } }, "instructions": "# Repository guardrails for coding agents\n\nThis project processes potentially identifying human genomic data. Automated contributors\nmust preserve the following boundaries:\n\n1. Never add patient data, genomic run files, credentials, direct identifiers or reports to\n Git. Use synthetic fixtures only.\n2. Treat every ISCN string as an expert-reviewable proposal. Do not remove the research-use\n warning or introduce automatic clinical release.\n3. Keep assay-specific QC and reportability rules versioned. A caller default is not a\n validated clinical threshold.\n4. Any change that can alter biol", "cost": { "context_tokens": 424 } }

ontseq-platform

About

Repository guardrails for coding agents

Use it

Install ontseq-platform and it becomes one more candidate the router can pick - when it fits.

Manifest

Reviews