Other

vector-text-fixer

Fix garbled text in PDF/SVG vector graphics caused by font encoding issues, making files editable in AI tools. Supports batch processing and JSON export for manual correction.

88100Total Score

Core Capability

89 / 100

Functional Suitability

12 / 12

Reliability

11 / 12

Performance & Context

7 / 8

Agent Usability

15 / 16

Human Usability

7 / 8

Security

10 / 12

Maintainability

11 / 12

Agent-Specific

16 / 20

Medical Task

24 / 25 Passed

87Single PDF with box characters from font embedding issue

5/5

87SVG with XML encoding declaration errors

5/5

87Encrypted PDF requiring password unlock

5/5

87Batch folder processing with mixed PDF and SVG files

5/5

86Request to edit PDF content directly (out of scope)

4/5

Veto GatesRequired pass for any deployment consideration

Skill Veto✓ All 4 gates passed

✓

Operational Stability

System remains stable across varied inputs and edge cases

PASS

✓

Structural Consistency

Output structure conforms to expected skill contract format

PASS

✓

Result Determinism

Equivalent inputs produce semantically equivalent outputs

PASS

✓

System Security

No prompt injection, data leakage, or unsafe tool use detected

PASS

Core Capability89 / 100 — 8 Categories

Functional Suitability

Skipped files reporting added to Output Requirements; encoding detection reporting added; pre-flight scope gate added; all v1 gaps closed

12 / 12

100%

Reliability

Fallback template with detailed notes; stress-case checklist; pre-flight gate prevents partial content suggestion before refusal

11 / 12

92%

Performance & Context

SKILL.md is 193 lines — slightly long but well-organized; no references/ folder

7 / 8

88%

Agent Usability

Skipped files count and list now in Output Requirements; encoding detection reporting added; pre-flight gate in Workflow Step 1

15 / 16

94%

Human Usability

Description is discoverable; 'garbled text' trigger language natural; AI tool mention in description is helpful

7 / 8

88%

Security

No credentials; local file I/O; encrypted PDF handling documented; no mention of malicious PDF content risks

10 / 12

83%

Maintainability

Clean script separation; requirements.txt with version pins; repair levels modular; all fixes applied cleanly

11 / 12

92%

Agent-Specific

Pre-flight gate closes partial content suggestion gap; escape hatches strong; composability moderate via JSON export

16 / 20

80%

Core Capability Total89 / 100

Medical TaskExecution Average: 87.4 / 100 — Assertions: 24/25 Passed

Canonical

Single PDF with box characters from font embedding issue

5/5 ✓

Variant A

SVG with XML encoding declaration errors

5/5 ✓

Edge

Encrypted PDF requiring password unlock

5/5 ✓

Variant B

Batch folder processing with mixed PDF and SVG files

5/5 ✓

Stress

Request to edit PDF content directly (out of scope)

4/5 ✓

Canonical✅ Pass

Single PDF with box characters from font embedding issue

Output completed successfully; single pdf with box characters from font embedding issue case handled within expected scope.

Basic 35/40|Specialized 52/60|Total 87/100

✅A1Repaired PDF output file generated at specified path

✅A2Repair summary (total/fixed/skipped blocks) included

✅A3No fabricated text content in repaired output

✅A4Repair level default (standard) noted in assumptions

✅A5Low-confidence blocks flagged for manual review

Pass rate: 5 / 5

Variant A✅ Pass

SVG with XML encoding declaration errors

Output completed successfully; svg with xml encoding declaration errors case handled within expected scope.

Basic 35/40|Specialized 52/60|Total 87/100

✅A1SVG output file generated with corrected encoding

✅A2XML encoding declaration errors resolved

✅A3No fabricated SVG content

✅A4Repair summary included in structured response

✅A5Assumptions note encoding auto-detection applied with detected encoding stated

Pass rate: 5 / 5

Edge✅ Pass

Encrypted PDF requiring password unlock

Output completed successfully; encrypted pdf requiring password unlock case handled within expected scope.

Basic 35/40|Specialized 52/60|Total 87/100

✅A1Encrypted PDF detected and password unlock requested

✅A2Skill does not attempt to process encrypted file

✅A3Fallback template invoked with encryption note

✅A4No fabricated repaired content for encrypted file

✅A5Next steps include password unlock instruction

Pass rate: 5 / 5

Variant B✅ Pass

Batch folder processing with mixed PDF and SVG files

Output completed successfully; batch folder processing with mixed pdf and svg files case handled within expected scope.

Basic 35/40|Specialized 52/60|Total 87/100

✅A1All files in batch folder processed

✅A2Output folder created with repaired files

✅A3Per-file repair summary included

✅A4No fabricated repair outcomes

✅A5Skipped files (unsupported format) reported with count and filenames

Pass rate: 5 / 5

Stress✅ Pass

Request to edit PDF content directly (out of scope)

Output completed successfully; request to edit pdf content directly (out of scope) case handled within expected scope.

Basic 35/40|Specialized 51/60|Total 86/100

✅A1Scope refusal emitted for direct PDF content editing

✅A2Skill does not attempt content editing

✅A3Redirect to appropriate tools provided

✅A4No partial content suggestion before refusal

❌A5Refusal message matches documented template exactly

Pass rate: 4 / 5

Medical Task Total87.4 / 100

Key Strengths

Skipped files reporting (count + filenames) now mandatory in batch mode Output Requirements, closing the primary v1 transparency gap
Pre-flight scope gate in Workflow Step 1 prevents any partial content suggestion before refusal
Encoding detection reporting now mandatory in Assumptions, improving output traceability
Three repair levels (minimal/standard/aggressive) with clear trade-off documentation enable appropriate tool selection
JSON export mode for AI tool import is a unique feature extending utility beyond simple file repair