omni-inference

The core OpenAI-compatible inference endpoints: chat completions, embeddings, images, audio (TTS/STT), moderations, rerank, and the Responses API. The primary integration surface for AI agents.

Quality

58%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/omni-inference/SKILL.md

Quality

Content

50%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 comprehensive, largely executable API reference, but it is padded with empty curl stubs and duplicates endpoint coverage across two generated halves. Restructuring into a lean overview with the bulk reference split out would materially improve it.

Suggestions

Remove the auto-generated endpoint-stub half (empty `-d '{}'` curl blocks) or merge it with the richer second half to eliminate duplication and recover token budget.

Move the per-endpoint curl reference into a separate bundled reference file (e.g. references/endpoints.md) and keep SKILL.md as a concise overview with one-level-deep links.

Add a short numbered workflow for the common case (authenticate -> discover models -> call endpoint -> handle errors) to give the reference an explicit sequence.

Dimension	Reasoning	Score
Conciseness	The first ~350 lines are auto-generated endpoint stubs each with an empty `-d '{}'` curl body that add little, and the same endpoints are re-covered in the second 'custom' half, making the body noticeably padded and duplicative.	2 / 5
Actionability	The second half provides fully executable, copy-paste curl examples with real payloads, model-discovery commands, and per-capability error codes for the common cases, though the stub half and OCR/proxy-subscription sections lack real payload examples.	4 / 5
Workflow Clarity	A consistent per-capability discover/call/errors pattern exists, but this is an API reference rather than a sequenced workflow, with no validation checkpoints or feedback loops, and the duplicated two-half structure muddies the flow.	3 / 5
Progressive Disclosure	There are clear section headers and signaled external references (OpenAPI spec, entry-point SKILL URL), but no local bundle files exist and the bulk API reference is inlined in a single monolithic file with the endpoints duplicated across two halves.	3 / 5
	Total	12 / 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 specific and third-person, clearly conveying the endpoint surface, but it omits any 'when to use' trigger guidance, which caps completeness. It is distinguishable but could be more trigger-oriented.

Suggestions

Add a 'Use when ...' clause naming concrete trigger phrases (e.g. 'Use when integrating AI agents that need chat completions, embeddings, TTS/STT, or reranking via an OpenAI-compatible API').

Consider leading with the concrete actions the skill enables rather than only the endpoint noun list to lift specificity from 4 to 5.

Dimension	Reasoning	Score
Specificity	Enumerates seven concrete endpoint categories ('chat completions, embeddings, images, audio (TTS/STT), moderations, rerank, and the Responses API'), giving comprehensive coverage of the capability surface, but they are named as endpoint nouns rather than concrete verb-actions.	4 / 5
Completeness	Clearly states 'what' (the inference endpoints) but has no 'Use when...' or equivalent explicit trigger guidance, so completeness is capped at 3 per the rubric guideline.	3 / 5
Trigger Term Quality	Includes natural terms a user would say ('chat completions', 'embeddings', 'images', 'TTS/STT', 'rerank', 'Responses API') with good coverage, though a few common synonyms are absent.	4 / 5
Distinctiveness Conflict Risk	The 'OpenAI-compatible inference endpoints' framing plus the specific endpoint list carves a mostly distinct niche with only minor overlap risk against related management skills.	4 / 5
	Total	15 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (671 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: diegosouzapw/OmniRoute
Path: skills/omni-inference/SKILL.md
Commit: ea0cdc5

Reviewed: about 6 hours 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.