using-model-endpoint

Call a registered model endpoint over its native HTTP API from the endpoint's scoped inference kernel (BASE_URL preloaded). Load once a task needs predictions from a registered model endpoint.

Quality

66%

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 ./configs/microservice/bff-service/configs/agent-skills/claude-science/using-model-endpoint/SKILL.md

Quality

Content

75%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 an efficient, well-structured overview that adds only non-obvious operational detail and cleanly routes the reader to sibling skills for lifecycle concerns. Its main gap is the absence of an executable code example for the core HTTP call.

Suggestions

Add a minimal copy-pasteable httpx/requests snippet showing a real request to f'{BASE_URL}/...' with the Bearer auth header so the core action is fully executable.

Show the exact compute_provider({'provider': '<slug>', 'code': '...'}) invocation as a concrete example rather than only describing its shape.

Optionally include a one-line validation tip (e.g. confirm a 200/expected schema before consuming the response) to strengthen the request/response workflow.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence — it states only non-obvious operational facts (BASE_URL preloading, the compute_provider slug, proxy requirements, no job lifecycle) with no padding or explanation of basic concepts.	5 / 5
Actionability	It gives concrete pointers (compute_provider call shape, Authorization: Bearer header, trust_env=False warning) but no copy-paste executable code example showing an actual httpx/requests call to BASE_URL, leaving the core action as guidance rather than runnable code.	3 / 5
Workflow Clarity	The single request/response flow is unambiguous with clear sequencing (build URL from BASE_URL, set auth, ride the proxy, direct request/response), and it explicitly notes the destructive/lifecycle operations belong elsewhere; it is a simple skill so does not require multi-step checkpoints.	4 / 5
Progressive Disclosure	It is a short, well-organized single file that clearly signals where related detail lives (provider runbook via skillName, the managed-model-endpoints skill), keeping references one level deep; there are no bundle files, so structure is appropriate.	4 / 5
	Total	16 / 20 Passed

Description

57%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 specific and distinguishes the skill's niche well, but it lacks an explicit 'Use when...' trigger clause and leans on technical jargon over natural user phrasing. This makes it moderately complete and distinct but not maximally triggerable.

Suggestions

Add an explicit 'Use when...' clause naming natural trigger phrases a user would say, e.g. 'Use when a task needs predictions from a registered model endpoint or when the user asks to call/query a deployed model.'

Soften jargon by including natural synonyms such as 'deployed model', 'model API', or 'inference endpoint' so the description triggers on user-natural language.

Consider listing the concrete actions available (e.g. 'Send HTTP requests, pass auth credentials, route through the sandbox proxy') to lift specificity toward comprehensive coverage.

Dimension	Reasoning	Score
Specificity	Names the concrete action ('Call a registered model endpoint over its native HTTP API') and the execution context ('from the endpoint's scoped inference kernel (BASE_URL preloaded)'), giving several specific operational specifics, though it stops short of listing multiple discrete actions.	4 / 5
Completeness	It clearly states what the skill does, but the 'when' is only weakly implied by 'Load once a task needs predictions from a registered model endpoint' — there is no explicit 'Use when...' trigger clause, which caps completeness at 3 per the guidelines.	3 / 5
Trigger Term Quality	It uses domain keywords like 'model endpoint', 'native HTTP API', 'inference kernel', and 'predictions', but these are technical jargon rather than the natural phrases a user would say ('I need to call a model'), and it omits common synonyms or file extensions.	3 / 5
Distinctiveness Conflict Risk	The 'registered model endpoint' / 'scoped inference kernel' framing carves a fairly distinct niche that is unlikely to fire for unrelated skills, with only minor overlap risk against the closely related managed-model-endpoints skill it itself references.	4 / 5
	Total	14 / 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: UnicomAI/wanwu
Path: configs/microservice/bff-service/configs/agent-skills/claude-science/using-model-endpoint/SKILL.md
Commit: 030c20b

Reviewed: 12 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.