deepdetect-pytorch-worker

Use when porting external PyTorch object detection models in DeepDetect through the external PyTorch worker backend.

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 ./bindings/python/deepdetect/pytorch_worker/SKILL.md

Quality

Content

81%Weight 40%Scale 1-5

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

The body is a well-structured, actionable porting workflow with executable commands, explicit validation, and error-recovery feedback loops; its main limitation is that all detail lives inline in a single file with no progressive disclosure to external reference materials.

Suggestions

Consider moving the detailed CLI smoke-test command block and the triage failure catalog into a referenced reference file (e.g. references/adapter-operations.md) so SKILL.md stays a lean overview with one-level-deep links.

Optionally extract the adapter implementation checklist (the 'Implement only the target-specific pieces' list) into a referenced template to improve progressive disclosure.

Trim a few explanatory list items in 'Build The Adapter' that restate what the surrounding bullets already imply.

Dimension	Reasoning	Score
Conciseness	The body is dense and assumes domain competence (no padding about what PyTorch or object detection is), and the conventions it does explain (xyxy vs cxcywh, one-based vs zero-based labels) are non-obvious project specifics; a few list explanations could be trimmed, matching 'efficient; minor instances of over-explanation'.	4 / 5
Actionability	It provides copy-paste-ready executable commands (rg terms, the pytest invocation, and the full CLI smoke-test command with concrete flags like --nclasses, --gpu --gpuid, --visdom), but the adapter Python implementation is given as a list of pieces to implement rather than shown, leaving minor gaps.	4 / 5
Workflow Clarity	A clear sequenced progression (Ground The Work -> Keep The Boundary -> Build The Adapter -> Add Generic Core Changes -> Validate -> Triage Failures) is paired with an explicit validation section and concrete error-recovery feedback loops in 'Triage Failures', matching the anchor for explicit validation steps plus feedback loops.	5 / 5
Progressive Disclosure	No bundle files exist and all content lives inline in SKILL.md, but it is well-organized into clearly headed sections with easy in-file navigation and no nested references; because the skill exceeds the under-50-line simple-skill exception and uses no one-level-deep external references, it sits at 'good structure, minor organization gaps' rather than the split-file anchor at 5.	4 / 5
	Total	17 / 20 Passed

Description

66%Weight 40%Scale 1-5

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 well-targeted to a specific niche with an explicit 'Use when' trigger and clear distinctiveness, but it lists only one action and offers limited keyword variation, leaving specificity and trigger coverage at mid-range.

Suggestions

Add 1-2 more concrete actions (e.g. 'import, adapt, and test') to broaden specificity beyond the single verb 'porting'.

Include natural synonyms or variations a user might say (e.g. 'integrate', 'adapt', 'wire up') to improve trigger term coverage.

Give the 'when' clause distinct user-facing scenarios (e.g. 'or when the user asks to import an external detector into DeepDetect') rather than restating the 'what'.

Dimension	Reasoning	Score
Specificity	The description names a concrete domain ('external PyTorch object detection models in DeepDetect') and one concrete action ('porting ... through the external PyTorch worker backend'), but lists only a single action rather than several, so it sits at the '1-2 concrete actions, not comprehensive' anchor rather than the multi-action anchor above.	3 / 5
Completeness	It has an explicit 'Use when ...' trigger (the 'when') and states what it does ('porting external PyTorch object detection models ... through the external PyTorch worker backend'), so both are present; the 'when' is a single trigger that largely mirrors the 'what' and lacks distinct user-mentionable scenarios, so it is not quite the comprehensive dual-trigger anchor at 5.	4 / 5
Trigger Term Quality	Relevant niche keywords are present ('porting', 'PyTorch object detection', 'DeepDetect', 'external PyTorch worker backend'), but the phrasing is fairly technical with no synonyms or variations (e.g. import, adapt) a user might also say, matching 'some relevant keywords but missing common variations'.	3 / 5
Distinctiveness Conflict Risk	The combination of 'external PyTorch worker backend', 'object detection', and 'DeepDetect' carves out a clear niche with distinct triggers and minimal realistic overlap with other skills.	5 / 5
	Total	15 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: jolibrain/deepdetect
Path: bindings/python/deepdetect/pytorch_worker/SKILL.md
Commit: 60527bd

Reviewed: 8 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.