launch-openshell-gator

Launch and supervise OpenShell gator agents. Use when starting gator on issues or PRs, checking gator sandboxes, building the gator sandbox image, restarting stuck gators, inspecting gator logs, or experimenting with gator harness/model overrides. Trigger keywords - launch gator, start gator, run gator, gator sandbox, supervised gator, gator logs, restart gator.

Quality

88%

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.

The content is highly actionable with a well-sequenced, validated workflow and clear structure. The main drag on token efficiency is heavy repetition of the input-normalization guard blocks across every code example, and the long single-file body does not progressively disclose detail into separate reference files.

Suggestions

Factor the repeated gateway-name / PR-number / sandbox-name regex validation into a single normalized snippet referenced once, then omit it from each launch block to cut substantial repeated tokens.

Consider moving the sentinel reference table and troubleshooting detail into a separate reference file (e.g. references/monitoring.md) and linking to it from the Monitoring section to improve progressive disclosure.

Tighten the long prose paragraph describing the launcher's internal provisioning steps, keeping only what the operator needs to act on.

Dimension	Reasoning	Score
Conciseness	The body is dense and operational with no generic concept padding, but the identical three-line gateway/PR-number/sandbox-name regex validation guard is repeated verbatim across roughly ten code blocks, which is avoidable token waste that could be factored into a shared snippet.	3 / 5
Actionability	Concrete, copy-paste-ready commands throughout — real './scripts/agents/run.sh ... --watch' invocations with clearly marked placeholders and dedicated blocks for PR watchers, issues, pairs, authorized launches, model experiments, and restarts covering the common cases.	5 / 5
Workflow Clarity	A six-step numbered preflight checklist with explicit validation checkpoints, guarded destructive actions ('Before deleting, check that the sandbox is truly stale'), and a troubleshooting section providing error-recovery feedback loops meets the top anchor.	5 / 5
Progressive Disclosure	Section structure is clear and a Key Paths table signals one-level-deep repo references well, but the ~400-line body is a single monolithic file with no content split into bundle reference files, so organization is good rather than exemplary.	4 / 5
	Total	17 / 20 Passed

Description

96%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.

The description is specific, trigger-rich, and completely answers both what the skill does and when to use it, with natural synonym coverage. Its only weakness is minor overlap with closely related OpenShell skills on the sandbox-inspection and log-inspection triggers.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'launch and supervise', 'building the gator sandbox image', 'restarting stuck gators', 'inspecting gator logs', 'experimenting with gator harness/model overrides' — giving comprehensive coverage of the operator workflow.	5 / 5
Completeness	Explicitly answers both 'what' ('Launch and supervise OpenShell gator agents') and 'when' (a 'Use when...' clause enumerating concrete triggering situations), matching the anchor for clearly answering both.	5 / 5
Trigger Term Quality	An explicit 'Trigger keywords - launch gator, start gator, run gator, gator sandbox, supervised gator, gator logs, restart gator' list supplies natural synonyms an operator would say, reinforced by the 'Use when starting gator on issues or PRs...' phrasing.	5 / 5
Distinctiveness Conflict Risk	The 'gator' niche and gator-prefixed triggers are clearly distinct, but 'checking gator sandboxes' and 'inspecting gator logs' create minor overlap risk with sibling skills like debug-openshell-cluster and openshell-cli rather than minimal conflict risk.	4 / 5
	Total	19 / 20 Passed

Validation

75%

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 — 12 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning
relative_links	Relative link issues: 9 missing	Warning
referenced_paths_exist	Referenced path issues: 16 missing, 16 deeper-than-1-level	Warning

	Total	12 / 16 Passed

Repository: NVIDIA/OpenShell
Path: .agents/skills/launch-openshell-gator/SKILL.md
Commit: 292559c

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.