Agentic Tools
Tools and infrastructure for reliable AI-assisted ontology, schema, and knowledge-base curation workflows. These components form the building blocks of an agent harness — the control plane that makes AI-assisted curation reproducible and reviewable.
For a guide on how these tools fit together, see Build your agentic harness.
Validators
linkml-reference-validator
Validates whether supporting text in structured records is actually present in cited references, helping enforce evidence-backed curation.
- When to use: After an agent generates structured annotations with literature citations — catch hallucinated or misattributed evidence before review.
- Install:
pip install linkml-reference-validator - GitHub | Docs | PyPI
linkml-term-validator
Checks LinkML schemas and datasets for correct use of external ontologies and controlled terms, improving consistency for agent-generated outputs.
- When to use: When agents generate data that references ontology terms — ensures terms exist and are used correctly.
- Install:
pip install linkml-term-validator - GitHub | Docs | PyPI
Provenance
ai-blame
Extracts provenance and audit trails from agent execution traces, enabling line-level attribution and post-hoc review for AI-assisted edits.
- When to use: After AI-assisted editing sessions — understand which changes an agent made, when, and in what context.
- Install:
pip install ai-blame - GitHub | Docs | PyPI
Agent Skills
curation-skills
Reusable skill packs for ontology and biocuration tasks, designed to make agent behavior more consistent, transparent, and domain-aware.
- When to use: When configuring agents for curation work — provide structured, domain-specific instructions instead of ad-hoc prompts.
- GitHub | Skills Article
MCP Servers
noctua-mcp
MCP server wrapping GO-CAM editing capabilities, enabling agentic interaction with Noctua/Barista workflows through a standardized interface.
- When to use: When agents need to create or edit GO-CAM models — provides structured access to the Noctua API.
- Install:
pip install noctua-mcp - GitHub | PyPI | noctua-py
oak-mcp
MCP server for ontology operations via the Ontology Access Kit (OAK), enabling agents to search, query, and traverse ontologies programmatically.
- When to use: When agents need to look up ontology terms, traverse hierarchies, or perform ontology operations during curation tasks.
- GitHub
System Instructions
CLAUDE.md and AGENTS.md
Files checked into the root of your repository that tell agents what the
repository is, which files are editable, and what conventions to follow.
Different agents read different files: Claude Code reads CLAUDE.md, Codex
reads AGENTS.md, and GitHub Copilot reads .github/copilot-instructions.md.
Keep one of them authoritative and make the others pointers to it. See One source of instructions.
- When to use: Always. Every repository that uses AI agents should have system instructions.
- See Instruct the GitHub agent
- Examples: CLAUDE.md in Mondo, CLAUDE.md in Uberon
Further Reading
- Case studies — these tools in the repositories that use them
- Patterns — the practices they support
- Build your agentic harness — how these tools compose into a harness
- Browse agentic tooling on GitHub
- ai4curation GitHub org