CtrlK
BlogDocsLog inGet started
Tessl Logo

tri-ai-collaboration

Complete development workflows where Claude writes the code while Gemini and Codex provide research, planning, reviews, and different perspectives. Claude remains the main developer. Use for complex projects requiring expert planning and multi-perspective reviews.

56

Quality

65%

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 ./.claude/skills/tri-ai-collaboration/SKILL.md

The canonical home for this skill is tri-ai-collaboration in fernandezbaptiste/Skrillz

SKILL.md
Quality
Evals
Security

Quality

Content

46%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.

Highly actionable and clearly sequenced, but severely bloated with redundant restatements and three near-duplicate end-to-end examples, all inlined in a single monolithic file with no progressive disclosure or validation checkpoints for destructive batch operations.

Suggestions

Collapse the three full end-to-end examples (OAuth, notifications, SaaS dashboard) into a single canonical worked example and move the others into a reference file, cutting the body by roughly half to eliminate the redundant 7-phase repetition.

Split the decision matrix, advanced collaboration patterns, and cost-optimization sections into one-level-deep reference files (e.g. references/patterns.md, references/decision-matrix.md) referenced from a lean SKILL.md overview.

Add explicit validation/verification checkpoints after each batch or destructive phase — verify generated code compiles/tests before proceeding past Phase 3, and require a review/confirmation gate before the Phase 7 auto-commit and deployment steps.

DimensionReasoningScore

Conciseness

The ~1160-line body restates the role model three times (Role Clarity, Power of Three, Summary) and includes three near-duplicate end-to-end examples that repeat the same 7-phase pattern, constituting substantial padding.

2 / 5

Actionability

Densely packed with concrete, copy-paste-ready bash commands, specific model names, and a task-to-AI decision matrix, but several JS examples contain placeholders ('// ... rest of improved implementation', '// Claude's implementation').

4 / 5

Workflow Clarity

Sequences are clearly numbered (Collaborative Loop, Phases 1-7), but validation is largely missing for batch/destructive steps — Phase 3 parallel code-gen just waits with no verification and Phase 7 auto-commits/deploys without checks, invoking the cap at 3.

3 / 5

Progressive Disclosure

No reference files exist; the entire 1160-line body is inlined in SKILL.md including three full multi-phase examples that clearly belong in separate reference files.

2 / 5

Total

11

/

20

Passed

Description

83%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, third-person description that clearly answers both what and when with concrete trigger language. Its main limitation is that the listed actions are somewhat high-level and the trigger clause lacks synonym coverage.

DimensionReasoningScore

Specificity

Names the domain and several concrete actions ('writes the code', 'research, planning, reviews, and different perspectives'), but the actions stay high-level rather than enumerating concrete per-AI verbs.

4 / 5

Completeness

Explicitly states both what the skill does (Claude writes code while Gemini/Codex advise) and when to use it ('Use for complex projects...'), with concrete trigger phrasing in third person.

5 / 5

Trigger Term Quality

The 'Use for complex projects requiring expert planning and multi-perspective reviews' clause supplies several natural terms a user would say, though it lacks common synonyms like 'second opinion' or 'code review'.

4 / 5

Distinctiveness Conflict Risk

The tri-AI-advisor framing is a distinct niche, but it self-references overlapping related skills (codex-cli, gemini-cli) and could trigger for generic code-review requests.

4 / 5

Total

17

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (1167 lines); consider splitting into references/ and linking

Warning

Total

15

/

16

Passed

Repository
fernandezbaptiste/Skrillz
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.