torch-geometric

Graph Neural Networks (PyG). Node/graph classification, link prediction, GCN, GAT, GraphSAGE, heterogeneous graphs, molecular property prediction, for geometric deep learning.

Quality

73%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./backend/cli/skills/coding/torch_geometric/SKILL.md

Quality

Content

65%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with executable code throughout, but it is verbose for a skill overview and lacks explicit validation checkpoints. Progressive disclosure is weakened by broken/incorrect reference paths.

Suggestions

Fix or remove the broken references: references/api_patterns.md does not exist and references/layer_capabilities.md should point to the actual references/layers_reference.md.

Move the full model definitions and complete training loops into reference files, keeping SKILL.md a lean overview with concise representative snippets.

Add explicit validation/verification checkpoints (e.g., assert dataset shapes, validate edge_index is undirected before training) in the training and custom-dataset workflows.

Dimension	Reasoning	Score
Conciseness	It assumes Claude's competence (no basic-concept padding), but the ~665-line body with many full model definitions and complete training loops is more than a lean overview needs and could be tightened by moving detail to reference files.	2 / 3
Actionability	Provides fully executable, copy-paste-ready code with concrete imports and parameters (install commands, full GCN/GAT/GraphSAGE classes, training loops, NeighborLoader, HeteroData).	3 / 3
Workflow Clarity	Sequences are present (numbered message-passing steps, train→eval loops), but validation checkpoints and feedback loops are missing or implicit for custom dataset processing and training.	2 / 3
Progressive Disclosure	Structure uses one-level-deep reference files and scripts, but two referenced paths are broken/non-existent (references/api_patterns.md, references/layer_capabilities.md) and naming is inconsistent with the actual layers_reference.md, weakening navigation.	2 / 3
	Total	9 / 12 Passed

Description

82%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, distinctive, and rich in natural trigger terms, but it omits an explicit 'Use when...' trigger clause, which caps its completeness. Adding a brief when-to-use clause would make it fully answer both what and when.

Suggestions

Add an explicit 'Use when...' clause, e.g. 'Use when working with graph-structured data, GNNs, or geometric deep learning.'

Confirm third-person voice is preserved (it is) and keep the action list as concise as it currently is.

Dimension	Reasoning	Score
Specificity	Lists multiple specific concrete capabilities and named architectures ("Node/graph classification, link prediction, GCN, GAT, GraphSAGE, heterogeneous graphs, molecular property prediction"), matching the score-3 anchor.	3 / 3
Completeness	The 'what' is well covered, but there is no 'Use when...' clause or equivalent explicit trigger, so per the guideline completeness is capped at 2 (when is only implied).	2 / 3
Trigger Term Quality	Includes natural terms users would say for GNN work (graph classification, link prediction, GCN, GAT, GraphSAGE, molecular property prediction), giving good coverage.	3 / 3
Distinctiveness Conflict Risk	"Graph Neural Networks (PyG)" / "geometric deep learning" is a clear, distinctive niche with specific triggers unlikely to conflict with other skills.	3 / 3
	Total	11 / 12 Passed

Validation

75%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (680 lines); consider splitting into references/ and linking	Warning
metadata_version	'metadata.version' is missing	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
referenced_paths_exist	Referenced path issues: 2 missing	Warning

	Total	12 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/coding/torch_geometric/SKILL.md
Commit: dc00e07

Reviewed: 10 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.