CtrlK
BlogDocsLog inGet started
Tessl Logo

veomni-debug

Use this skill for ANY bug, error, crash, wrong output, loss divergence, gradient explosion, test failure, CUDA error, distributed training hang, checkpoint load failure, or unexpected behavior. Covers both quick fixes (clear root cause) and complex debugging (unclear cause). Trigger: 'fix bug', 'fix error', 'broken', 'crash', 'doesn't work', 'fails with', 'loss NaN', 'training hangs', 'FSDP error', 'OOM'.

72

Quality

89%

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

88%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.

A well-structured, highly actionable debugging protocol with strong workflow sequencing and verification loops; the only gap is the absence of bundle files to offload the checklists and subagent template.

Suggestions

Move the Appendix verification-subagent prompt and the Domain-Specific Checklists into a reference file (e.g. references/verification.md) referenced one level deep from SKILL.md to tighten conciseness and lift progressive_disclosure.

Consider extracting Common Pitfalls to references/pitfalls.md so the main body stays focused on the decision path (Quick Path vs Full Protocol).

Trim the redundant TodoWrite example block — a single line instructing TodoWrite use would convey the same guidance in fewer tokens.

DimensionReasoningScore

Conciseness

Lean debugging protocol that assumes Claude's competence and avoids explaining basic concepts; the only trimmable block is the appendix verification-subagent template, keeping it just below the 5 anchor.

4 / 5

Actionability

Fully executable guidance: concrete commands ('git log --oneline -10', 'uv venv .venv-default', 'VIRTUAL_ENV=.venv-default uv sync --extra gpu --dev', 'pytest tests/<module>/'), specific file paths, and a copy-paste-ready subagent prompt.

5 / 5

Workflow Clarity

Phases 1–5 are clearly sequenced with explicit validation gates ('Verification gate' in Phase 3, verify step in Phase 4 and Quick Path), feedback loops ('If wrong, update understanding'), a three-strike escalation rule, and per-domain checklists.

5 / 5

Progressive Disclosure

No bundle files exist; the single SKILL.md is well-sectioned with clear headers and clearly-signaled one-level references (.agents/knowledge/constraints.md, architecture.md), but all reference material is inline so it sits at the 4 anchor rather than 5.

4 / 5

Total

18

/

20

Passed

Description

90%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, highly-triggerable description with comprehensive natural keywords and explicit what/when guidance, docked only for the second-person voice that the rubric penalizes on specificity.

DimensionReasoningScore

Specificity

Lists many concrete failure types ('loss divergence, gradient explosion, CUDA error, distributed training hang, checkpoint load failure') plus two action categories (quick fixes / complex debugging), but the second-person imperative 'Use this skill for…' triggers the −1 voice penalty, dropping the base-4 to 3.

3 / 5

Completeness

Explicitly answers both what ('Covers both quick fixes (clear root cause) and complex debugging (unclear cause)') and when ('Use this skill for ANY bug, error, crash…' with a dedicated Trigger clause), matching the 5 anchor.

5 / 5

Trigger Term Quality

Comprehensive natural triggers including synonyms and technical shorthand: 'fix bug', 'broken', 'crash', 'doesn't work', 'fails with', 'loss NaN', 'training hangs', 'FSDP error', 'OOM' — phrases a user would naturally say.

5 / 5

Distinctiveness Conflict Risk

Clear niche in distributed-training debugging (FSDP, MoE, sequence parallel, checkpoint load) with distinct triggers; minimal overlap risk with general skills.

5 / 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.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
ByteDance-Seed/VeOmni
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.