CtrlK
BlogDocsLog inGet started
Tessl Logo

v4-best-practices

Use when working with deepseek-v4-pro or deepseek-v4-flash in thinking mode on multi-step or plan-driven tasks. Provides rules to prevent stale references, unverified plan assumptions, and vague plan output.

76

Quality

95%

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

SKILL.md
Quality
Evals
Security

Quality

Content

100%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tight, well-structured set of three preventive rules: lean, concrete, and explicitly validation-oriented, with no bundle files needed for a sub-50-line instruction skill.

DimensionReasoningScore

Conciseness

The body is lean — a one-line intro plus three terse rules, each with a compact bad/good example and a single 'Failure avoided' line — with no padding or explanation of concepts Claude already knows, matching the 'every token earns its place' anchor 3.

3 / 3

Actionability

Each rule gives concrete, specific guidance naming exact tools and parameters (grep_files/read_file with a pattern, agent type='verifier' model='deepseek-v4-flash', path:line references) with clear bad/good contrasts; as an instruction-only skill its actionable guidance is not penalized for using tool-call syntax rather than literal code.

3 / 3

Workflow Clarity

Rules are explicitly sequenced (1-2-3) and each is itself an explicit validation checkpoint (confirm path exists, verifier confirms assumptions, confirmed path:line references), satisfying the 'clear sequence with explicit validation steps' anchor 3 better than anchor 2's 'checkpoints missing or implicit'.

3 / 3

Progressive Disclosure

The skill is under 50 lines with no external references needed and is organized into clear numbered sections, so per the simple-skills scoring note it qualifies for anchor 3 on well-organized structure alone.

3 / 3

Total

12

/

12

Passed

Description

90%

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 well-formed description with an explicit 'Use when' trigger, strong distinctiveness, and good natural trigger terms; its only weakness is that it states outcomes prevented rather than concrete actions performed, keeping specificity at 2.

Suggestions

Lead with concrete actions the skill performs (e.g., 'Verifies file/symbol references via grep before editing, spawns a flash verifier sub-agent before multi-file plans') so the description lists specific actions rather than only naming failure classes it prevents.

Consider naming the tool calls the rules prescribe (grep_files/read_file, verifier sub-agent) to make capabilities tangible at a glance.

DimensionReasoningScore

Specificity

The phrase 'Provides rules to prevent stale references, unverified plan assumptions, and vague plan output' names three specific failure classes, but describes what is prevented rather than listing concrete actions the skill performs (the actual rules live in the body), matching 'names domain and some actions, but not comprehensive' better than the multi-concrete-action anchor 3.

2 / 3

Completeness

It explicitly answers both 'when' ('Use when working with deepseek-v4-pro or deepseek-v4-flash in thinking mode on multi-step or plan-driven tasks') and 'what' ('Provides rules to prevent...'), matching the anchor 3 example with an explicit 'Use when' trigger.

3 / 3

Trigger Term Quality

'deepseek-v4-pro or deepseek-v4-flash in thinking mode on multi-step or plan-driven tasks' gives good natural-phrase coverage a user would say (model names, thinking mode, multi-step/plan-driven), matching anchor 3; it is not merely jargon.

3 / 3

Distinctiveness Conflict Risk

Tight niche scoped to specific V4 models in thinking mode on multi-step/plan-driven tasks, giving distinct triggers unlikely to fire for unrelated skills, matching anchor 3.

3 / 3

Total

11

/

12

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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
Hmbown/CodeWhale
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.