CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

think-first

Reflection first problem solving for Convex projects: root cause before code, two candidate fixes, a 98 percent confidence bar, and changes limited to the files the request needs. Use before implementing any fix or feature, when a bug report comes in, when a previous fix did not work, or when the request is vague.

Quality

94%

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

SKILL.md
Quality
Evals
Security

Quality

Content

96%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.

An exceptionally tight, actionable instruction skill: concrete rules, a gated decision workflow, and dual checklists leave no ambiguity about what to do before and after a change. The only structural improvement is moving the Convex docs link catalog into a references file now that the body has grown past simple-skill size.

DimensionReasoningScore

Conciseness

The body is lean and imperative throughout ('Slow down for thirty seconds before the first edit. It saves an hour later.'; 'Never assume. If anything is unclear, ask one focused question and stop.') with no explanations of concepts Claude already knows; even asides like 'It does not prevent write conflicts' carry decision weight, matching the anchor 5 'every token earns its place' example.

5 / 5

Actionability

For an instruction-only skill the guidance is fully concrete and executable: 'compare its owner to the caller from `ctx.auth`', 'Run `git log --date=short`', and an explicit prohibition list ('never run `reset --hard`, `checkout -- .`, `clean -fd`, or `stash drop`'), matching anchor 5's copy-paste-ready specificity.

5 / 5

Workflow Clarity

The five-step reflect sequence is clearly gated ('Do not write code below 98 percent confidence... Below that: read more, ask, or propose a plan instead of a diff') and bracketed by two checklists with verification items ('The fix was verified, not just written'), matching anchor 5's explicit validation steps, feedback loops, and checklists.

5 / 5

Progressive Disclosure

Sections are well-organized with no nested references, but at ~98 lines (above the under-50-line simple-skill exception) the 14-line Convex docs link dump is content that could live in a references/ file, and the assets/ logo files are never referenced from the body — minor organization gaps matching anchor 4 rather than anchor 5's 'content appropriately split'.

4 / 5

Total

19

/

20

Passed

Description

88%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.

A strong description that clearly states what the skill does and when to use it, with concrete and largely idiosyncratic behavioral commitments. The main gaps are missing common trigger synonyms and a couple of overly broad 'when' clauses that could cause overlap with general coding-workflow skills.

Suggestions

Add natural trigger synonyms such as 'error', 'issue', 'debug', or 'broken' to the 'Use when' clause so it fires on the ways users actually report problems.

Tighten the broadest triggers ('before implementing any fix or feature', 'when the request is vague') toward the Convex-specific context to reduce overlap with general planning skills.

DimensionReasoningScore

Specificity

The colon expansion enumerates four concrete, measurable behaviors — 'root cause before code, two candidate fixes, a 98 percent confidence bar, and changes limited to the files the request needs' — matching the comprehensive multi-action anchor rather than anchor 4, which requires minor coverage gaps.

5 / 5

Completeness

It explicitly answers both 'what' (the enumerated reflection-first behaviors) and 'when' with concrete trigger phrases ('Use before implementing any fix or feature, when a bug report comes in, when a previous fix did not work, or when the request is vague'), mirroring the anchor 5 exemplar structure.

5 / 5

Trigger Term Quality

Natural phrases like 'bug report', 'previous fix did not work', and 'request is vague' are present, but common synonyms users would say ('error', 'issue', 'debug', 'broken') are missing, fitting anchor 4's 'good keyword coverage; a few natural terms missing' rather than anchor 5's synonym-and-extension comprehensiveness.

4 / 5

Distinctiveness Conflict Risk

The Convex scoping and idiosyncratic behaviors (98 percent confidence bar, two candidate fixes) create a distinct niche, but broad triggers like 'before implementing any fix or feature' and 'request is vague' risk overlap with general planning or coding-workflow skills, matching anchor 4 rather than anchor 5's minimal conflict risk.

4 / 5

Total

18

/

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

waynesutton/convexskills

Path

skills/think-first/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.