20/ 100Poor
2 critical
6 warnings
10 suggestions
109 checks passed
Domain scores
Anti-Slop
80
Accessibility
100
Enterprise
100
Issues
Critical (2)
1.1-name-charsline 2
Name must be lowercase letters, numbers, and hyphens only
Fix: Use lowercase letters, numbers, and hyphens
7.2-injection-riskline 130
Potential injection risk: unescaped variable in shell
Fix: Sanitize inputs and use parameterized commands
Warnings (6)
5.1-no-style-sectionline 1
No writing style section found
Fix: Add a '## Writing Style' or '## Anti-Slop' section
16.2-no-guardrailsline 11
Skill uses Bash or destructive operations without safety guardrails
Fix: Add guardrails: specify forbidden commands, require confirmation for destructive actions, or use --dry-run patterns
15.1-no-budget-capline 11
Multi-agent skill has no budget or turn cap; runaway costs are possible
Fix: Add max_turns, cost_limit, or similar budget constraint for subagents
4.5-desc-no-triggersline 2
Description lacks triggering conditions
Fix: Add phrases like 'Use when...', 'Triggers on...', or 'Applies to...' to help Claude select this skill
1.2-desc-whenline 3
Description missing WHEN clause - recommended by agentskills spec
Fix: Consider adding 'Use when...' or 'Activate when...' to description
12.5-no-exit-criterialine 15
4 of 4 phases lack exit criteria; Claude may skip phases or run past completion
Fix: Add exit criteria to each phase: 'Exit when: [specific condition]'
Suggestions (10)
5.2-no-do-dontline 1
Missing DO/DON'T lists for style guidance
Fix: Add explicit DO and DON'T lists for writing patterns
16.3-no-rollbackline 11
State-modifying skill lacks rollback or recovery guidance
Fix: Add rollback instructions: how to undo changes if something goes wrong (e.g., git revert, backup/restore steps)
16.4-no-idempotencyline 11
Skill does not document whether it is safe to re-run
Fix: Add idempotency guidance: state whether running the skill twice produces the same result or causes side effects
17.1-no-input-contractline 11
No input contract found; unclear what data the skill expects
Fix: Document what the skill expects as input (format, required fields, constraints)
17.3-no-mode-awarenessline 11
Skill does not clarify standalone vs composed behavior
Fix: Clarify whether the skill works standalone, in pipelines, or both
3.3-name-single-wordline 2
Single-word names are less discoverable
Fix: Consider a two-word name like 'task-tracker' or 'code-reviewer'
18.1-no-confidence-signalline 1
No confidence signal found for orchestrator decision-making
Fix: Add confidence indicators (e.g., confidence: high/medium/low) so orchestrators can route or retry
15.6-subagent-no-return-formatline 141
Subagent spawning at line 141 lacks return format specification
Fix: Specify what format the subagent should return results in (JSON, markdown table, structured text)
4.3-output-formatline 1
Skill mentions output but doesn't specify format
Fix: Add an '## Output' section with format examples
1.19-non-portable-fieldline 4
Frontmatter keys outside the agentskills spec: argument-hint. Claude Code loads the skill with them; claude.ai uploads and the Skills API reject it with 'Unexpected key(s)'
Fix: Nothing to do for Claude Code or plugin distribution. To upload to claude.ai or the Skills API, ship frontmatter with only name, description, license, compatibility, metadata and allowed-tools
OWASP Agentic (Cat 26)
Agent surface detected: declares tool beyond Read/Glob/Grep: Bash; declares tool beyond Read/Glob/Grep: Task; performs consequential actions
2 warning
-
26.02-ASI02-unconstrained-bash
warning
Bash is granted with no constraint note (which commands, why)
Fix: document which Bash commands the skill runs and why, or narrow the grant
-
26.11-ASI11-no-audit
warning
skill performs consequential actions with zero logging/audit language
Fix: document a logging/audit/provenance step (correlation id, structured log, handoff) for consequential actions
ASI-01: skippedASI-02: failASI-03: skippedASI-04: skippedASI-05: skippedASI-06: skippedASI-07: skippedASI-08: skippedASI-09: skippedASI-10: skippedASI-11: fail
Badges
skillcheck
✕ passed
skillcheck
✕ antislop 100
skillcheck
WCAG AA
skillcheck
WCAG AAA
skillcheck
enterprise
Get a report like this for your own skill
Run the free validator to see your score in seconds. Pro adds the substance checks: anti-slop, security, accessibility and agent readiness.