file-validator

Validate file formats and content safely

Invalid

This skill can't be scored yet

Validation errors are blocking scoring. Review and fix them to unlock Quality, Impact and Security scores. See what needs fixing →

Evaluation results

100%

17%

Document Ingestion Security Validator

Path traversal prevention and extension whitelist

Criteria

Without context

With context

Python implementation

100%

Uses os.path.basename

66%

100%

Basename applied before checks

100%

Extension check present

100%

Allowed .txt extension

100%

Allowed .csv extension

100%

Allowed .json extension

100%

Allowed .md extension

100%

Rejects unsupported extension

100%

Path traversal rejected

86%

100%

Failure reason returned

100%

Valid result returns filename

100%

Without context: $0.2493 · 1m 12s · 15 turns · 16 in / 4,362 out tokens

With context: $0.6184 · 2m 7s · 34 turns · 1,085 in / 7,680 out tokens

98%

Upload Safety Checker for Document Portal

File size limit and MIME type detection

Criteria

Without context

With context

Python implementation

100%

Uses mimetypes module

100%

mimetypes.guess_type used

100%

Does NOT use python-magic

100%

Size check with os.path.getsize

100%

10MB size limit enforced

100%

Oversized file rejected

100%

MIME type in valid result

100%

valid field in result

30%

100%

Failure reason included

100%

80%

Without context: $0.3338 · 1m 21s · 20 turns · 20 in / 5,059 out tokens

With context: $0.5848 · 2m 5s · 34 turns · 279 in / 7,434 out tokens

80%

45%

Secure File Processing Pipeline for Compliance Reports

Directory restriction and structured validation results

Criteria

Without context

With context

Python implementation

100%

Uses expanduser for home dir

Safe directory is ~/documents/

Joins safe dir with basename

100%

Files outside directory rejected

100%

valid key in result

100%

reason key on failure

100%

filename key on success

100%

size key on success

100%

Results output as JSON

100%

Without context: $0.4315 · 1m 34s · 24 turns · 24 in / 6,057 out tokens

With context: $0.5546 · 2m 2s · 31 turns · 278 in / 7,078 out tokens

Repository: cisco-ai-defense/skill-scanner
Commit: c0a625f

Evaluated: 17 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Document Ingestion Security Validator Upload Safety Checker for Document Portal Secure File Processing Pipeline for Compliance Reports

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.