CtrlK
BlogDocsLog inGet started
Tessl Logo

agent-only-skill

An agent-only skill not intended for direct user invocation

30

Quality

23%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./tests/fixtures/sample-plugin/skills/agent-only-skill/SKILL.md

The canonical home for this skill is agent-only-skill in EveryInc/compound-engineering-plugin

SKILL.md
Quality
Evals
Security

Quality

Content

37%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 one-line placeholder with no instructions, examples, or workflow — it gives an executing agent nothing to act on. The conciseness is perfect only because the content is absent; actionability and workflow clarity fail at the floor. The skill needs its actual operational guidance written into the body (or pointed to via real reference files).

Suggestions

Write the actual instructions the agent should follow when this skill is invoked — concrete commands, code, or specific procedural guidance — since currently there is nothing actionable to execute.

If the skill involves multi-step or risky operations, sequence the steps explicitly and include a validation/verification checkpoint where the agent confirms the outcome before proceeding.

If detailed material belongs in separate files, create the references/ files and link them clearly from a short overview in the body, rather than leaving the body as a bare placeholder.

DimensionReasoningScore

Conciseness

The body ('Agent-only skill body.') is a single line with zero padding, no over-explanation, and no concepts Claude already knows — trivially lean; the skill's problem is emptiness, not verbosity, which is captured under actionability.

5 / 5

Actionability

The body provides no concrete code, commands, or specific direction of any kind — it is a placeholder sentence that describes rather than instructs, matching the 'entirely vague or abstract' anchor.

1 / 5

Workflow Clarity

No steps, sequence, or single unambiguous action is stated anywhere. The simple-skill exception does not apply because the one action the skill performs is never identified, so the 'steps missing or incoherent' anchor governs.

1 / 5

Progressive Disclosure

No bundle files exist (references/, scripts/, assets/ are absent) and the body has no buried or nested references, so nothing is misplaced — but there are also no sections or overview structure to navigate, and nothing signalled to defer to. It sits between 'minimal structure' and the simple-skill ideal.

3 / 5

Total

10

/

20

Passed

Description

8%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 a placeholder that classifies the skill's invocation mode without conveying any capability, purpose, or trigger context. Even allowing that agent-only skills are not user-triggered, it fails to tell an agent dispatcher what the skill does or when it applies. It needs at minimum a concrete statement of its function and its intended calling contexts.

Suggestions

State the concrete capability in third person, e.g. 'Reviews X and produces Y when invoked by an agent' — replace the 'not intended for direct user invocation' framing with a positive description of what it does.

Add explicit usage guidance for the intended audience (agents/harness), such as 'Use when the orchestrating agent needs to ...', so completeness covers both what and when.

Include domain-specific nouns or terms that distinguish this skill from other agent-only skills to reduce conflict risk.

DimensionReasoningScore

Specificity

The description 'An agent-only skill not intended for direct user invocation' names no concrete actions or capabilities whatsoever — it only states what the skill is not for, which matches the 'entirely vague; no concrete actions; pure abstract language' anchor.

1 / 5

Completeness

It offers only a vague 'what' ('An agent-only skill') and, while it explicitly rules out user invocation, it gives no 'when' — it never states when the skill should actually be used. This matches the 'vague what and no when' anchor; it cannot score 3 because there is no clear statement of what the skill does.

2 / 5

Trigger Term Quality

There are no natural keywords a user (or agent) would say when needing this skill; the text contains only generic classification language with no domain terms or synonyms, matching the lowest anchor.

1 / 5

Distinctiveness Conflict Risk

'An agent-only skill' is entirely generic and would be indistinguishable from any other agent-only skill, creating conflict with virtually any similarly-classified skill — the lowest anchor.

1 / 5

Total

5

/

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
crdant/compound-engineering-plugin
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.