A lean, highly actionable porting workflow with strong sequencing, explicit validation, and real error-recovery guidance. Its main gaps are minor: a missing worker.py code skeleton and slight repetition of the boundary rules.

Dimension	Reasoning	Score
Conciseness	The body is dense and imperative throughout (e.g. 'Prefer subclassing the existing detection training base', 'Raise typed worker errors') with no explanation of concepts Claude already knows; the only trimmable redundancy is the don't-commit-target-code rule stated in both 'Keep The Boundary' and 'Add Generic Core Changes Only'.	4 / 5
Actionability	Copy-paste-ready commands are provided (the pytest invocation and the full CLI smoke-test command with real flags) plus the adapter directory layout and concrete rg search terms, but the core deliverable, worker.py, is described as a checklist rather than shown with a code skeleton.	4 / 5
Workflow Clarity	A clear sequenced workflow (Ground The Work, Keep The Boundary, Build The Adapter, Generic Core Changes, Validate, Triage Failures) with explicit validation steps (focused tests, gated real-upstream tests, CLI smoke tests) and a dedicated feedback-loop section mapping import/eval/rendering failures to specific inspection steps.	5 / 5
Progressive Disclosure	A well-sectioned single-file skill with no bundle and no dangling references; at ~180 lines, some detail (full CLI examples, the triage matrix, the manifest field spec) could be split into reference files, which keeps it just short of the ideal split.	4 / 5
	Total	17 / 20 Passed

Description

73%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.

A tightly scoped, correctly triggered description that names a distinct niche (DeepDetect external PyTorch detection worker porting) but describes only a single action. It answers when to use the skill explicitly and adequately answers what, without padding or buzzwords.

Suggestions

State what the skill actually does beyond the trigger, e.g. 'Generates adapter scaffolding under extern/pytorch_workers/, handles box/label conversion, and validates with pytest and CLI smoke tests.'

Add action verbs that cover the workflow's scope (port, import, adapt, test) so the what-side lists several concrete actions instead of only 'porting'.

Include common synonyms such as 'import' or 'integrate external detection models' to broaden natural trigger coverage.

Dimension	Reasoning	Score
Specificity	The description names a concrete domain and mechanism ('porting external PyTorch object detection models in DeepDetect through the external PyTorch worker backend') but offers only one action verb, 'porting', with no mention of importing, adapting, testing, or generating adapters.	3 / 5
Completeness	An explicit 'Use when porting...' trigger clause is present, and the what ('porting... through the external PyTorch worker backend') is concrete, but the what-side is thin and largely circular with the when-clause, so it falls short of the fully explicit what+when anchor.	4 / 5
Trigger Term Quality	Terms like 'porting', 'PyTorch', 'object detection models', and 'DeepDetect' are exactly what a user would say, but natural synonyms such as 'import', 'integrate', or 'external model' are missing.	4 / 5
Distinctiveness Conflict Risk	'DeepDetect', 'external PyTorch worker backend', and 'object detection models' carve out a clear niche with essentially no overlap risk against other skills.	5 / 5
	Total	16 / 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: ff9edb3

Reviewed: 2 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.