aris-compute-guard

Mandatory pre-flight compute resource check before running experiments. Detects whether local/remote GPU or compute resources are actually available. If resources are unavailable, STOPS the experiment pipeline immediately and reports to the user — preventing the model from hallucinating fake experiment results. Use when: about to run experiments, deploy training, or any GPU-intensive task.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

81%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 highly actionable, well-sequenced guard skill with executable checks and an explicit decision gate. The main weakness is redundancy: the 'do not fabricate results' injunction is stated four times and could be tightened to improve token efficiency.

Suggestions

Consolidate the repeated anti-fabrication warning — it appears in the description, CRITICAL RULE, Step 4 message, and Rules; state it once authoritatively and reference it.

Move the per-environment detection snippets (Local CUDA, MPS, CPU-only, Remote SSH, Vast.ai, Modal) into a referenced file or fold the less-common ones (Vast.ai, Modal) into a compact list to reduce main-body length.

Replace the '<server>' placeholder in the SSH block with a note pointing to the CLAUDE.md remote-host field so the command is fully copy-pasteable.

Dimension	Reasoning	Score
Conciseness	Code blocks and thresholds are tight and free of basic-concept padding, but the anti-fabrication warning is repeated across the description, CRITICAL RULE, Step 4 template, and Rules section, which could be consolidated.	3 / 5
Actionability	Copy-paste-ready executable commands for every environment with concrete thresholds ('memory.used < 500 MiB', 'utilization >= memory.total * 0.9', 'ConnectTimeout=10'), a decision-gate table, and a fill-in report template.	5 / 5
Workflow Clarity	Clear five-step sequence with an explicit decision gate (Step 3), a stop-and-report branch (Step 4), and an explicit error-recovery rule ('If the check itself fails... treat it as unavailable and report').	5 / 5
Progressive Disclosure	Well-organized into clearly labeled sections with all content legitimately inline for a self-contained guard skill; no external bundle files are present, and the per-environment checks readably belong together, though the file is longer than the simple-skill exception threshold.	4 / 5
	Total	17 / 20 Passed

Description

92%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, concrete description that clearly states both capability and trigger conditions in third person. Trigger term coverage is good but could add a few more natural synonyms for experiment/training tasks.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Detects whether local/remote GPU or compute resources are actually available', 'STOPS the experiment pipeline immediately', 'reports to the user' — with comprehensive coverage of the guard's behavior.	5 / 5
Completeness	Explicitly answers both what (pre-flight compute check that detects availability, stops the pipeline, reports) and when ('Use when: about to run experiments, deploy training, or any GPU-intensive task') with concrete triggers.	5 / 5
Trigger Term Quality	Natural trigger phrases are present ('about to run experiments, deploy training, or any GPU-intensive task') but common synonyms like 'train a model', 'fine-tuning', or 'inference' are missing.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (mandatory pre-flight compute guard that halts the pipeline to prevent hallucinated results) with distinct triggers and minimal overlap with other skills.	5 / 5
	Total	19 / 20 Passed

Validation

87%

Warnings & errors only

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

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: OpenLAIR/dr-claw
Path: skills/aris-compute-guard/SKILL.md
Commit: 4b35756

Reviewed: 5 days ago

Table of Contents

Discovery Implementation Validation

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.