CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

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

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/omni-inference/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

60%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 delivers strong, executable API guidance with a consistent discover-example-errors structure and well-verified one-level-deep references. Its main weakness is token efficiency: a duplicated endpoint index, repeated auth boilerplate, and two conflicting authentication stories inflate the skill without adding capability.

Suggestions

Replace the ~140-line inline endpoint list with a short pointer to references/endpoints.md (which already contains the same table of contents with anchors), keeping only the handful of core inference endpoints inline.

State the OMNIROUTE_URL/OMNIROUTE_KEY setup once in the Authentication section and drop the seven repeated "Requires... See entry-point SKILL" lines, resolving the conflict with the Bearer-token/login instructions.

Add the Authorization header to the "Discover" curl examples so they are executable given the stated requirement that all requests carry a valid Bearer token.

DimensionReasoningScore

Conciseness

The ~140-item endpoint list duplicates the table of contents already present in references/endpoints.md, the Overview repeats the frontmatter description verbatim, and the line "Requires `OMNIROUTE_URL` and `OMNIROUTE_KEY`. See entry-point SKILL..." is repeated seven times across sections. The core capability sections are lean, so this is noticeably verbose rather than severely padded.

2 / 5

Actionability

Concrete, copy-paste-ready curl examples with headers, JSON bodies, response shapes, and error codes with remediation cover chat, images, TTS, STT, embeddings, web search, and web fetch. Minor gap: every "Discover" curl (e.g. `curl $OMNIROUTE_URL/v1/models | jq '.data[].id'`) omits the Authorization header that the Authentication section says all requests require, so those commands would fail as written.

4 / 5

Workflow Clarity

Each capability follows a clear, consistent sequence (setup requirement, endpoints, discover, example, errors) with error-recovery guidance such as "honor `Retry-After`" and "check `/v1/models`". Auth setup is deferred to an external GitHub URL rather than stated as an explicit first step, which leaves minor sequencing gaps.

4 / 5

Progressive Disclosure

All endpoint links resolve to real anchors in the bundled references/endpoints.md (verified: one level deep, clearly signaled, with per-endpoint detail properly offloaded). The inline 140-link list duplicates the reference file's own table of contents, and the external pointers (GitHub auth skill, OpenAPI spec) are not part of the bundle — minor organization gaps.

4 / 5

Total

14

/

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 distinctive, enumerating concrete inference capabilities with natural trigger keywords. Its main gap is the missing "Use when..." trigger guidance, which caps completeness and leaves the activation scenario implicit.

Suggestions

Add an explicit trigger clause, e.g. "Use when integrating an AI agent with OpenAI-compatible chat, embeddings, image, audio, or rerank endpoints."

Include common user phrasings like "text-to-speech", "transcription", and "speech-to-text" alongside the TTS/STT abbreviation to broaden trigger coverage.

DimensionReasoningScore

Specificity

Names several specific capabilities — "chat completions, embeddings, images, audio (TTS/STT), moderations, rerank, and the Responses API" — with comprehensive coverage of the inference domain. Not 5: these are endpoint categories rather than concrete actions, and capabilities present in the body (web search, web fetch) are omitted.

4 / 5

Completeness

The "what" is clear and specific, but there is no "Use when..." clause or equivalent trigger guidance — "The primary integration surface for AI agents" is a role statement, not a usage trigger. Per the rubric guideline, a missing trigger clause caps completeness at 3; it is not 2 because the "what" is fully explicit.

3 / 5

Trigger Term Quality

Good natural-keyword coverage: "chat completions", "embeddings", "TTS/STT", "OpenAI-compatible", "Responses API" are terms users would say when needing this skill. A few natural variants ("text-to-speech", "transcription", "speech-to-text") are missing, keeping it below 5.

4 / 5

Distinctiveness Conflict Risk

Clear niche (OpenAI-compatible inference endpoints) with named services that distinguish it from generic API skills; only minor overlap risk with closely related skills in the same family, such as the omni-auth entry-point skill the body references.

4 / 5

Total

15

/

20

Passed

Validation

100%

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

diegosouzapw/OmniRoute

Path

skills/omni-inference/SKILL.md

Reviewed

Table of Contents

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.