CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

ai-subtitle-remover

Remove hardcoded subtitles from user-selected videos, watermarks from TikTok and other supported short-video share links, or watermarks and unwanted text from selected images using the 550W Open API. Also check tasks and credits.

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./claude-plugin/550w-ai-subtitle-remover/skills/ai-subtitle-remover/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

75%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.

An exceptionally lean, well-structured instruction body with good checkpoint discipline (timeout re-check, delete confirmation, up-front billing). The dominant defect is that the referenced CLI entry point dist/550w-skill.cjs is missing from the bundle, which undermines the otherwise executable guidance, and batch image processing lacks an explicit verification loop.

Suggestions

Ship the bundled CLI at dist/550w-skill.cjs (or correct the path to wherever the 550w-skill.cjs script actually lives) — as delivered, the skill's only executable command points at a nonexistent file.

Add one inline example request for the primary workflow action (e.g. a minimal {"action":"workflow","params":{...}} payload) alongside the queryCredits example so the common case is copy-paste ready without opening the reference.

Spell out the per-image verification loop for batch watermark removal: after each image, check imageWatermarkTaskDetail status and only submit the next image once the previous result is confirmed.

DimensionReasoningScore

Conciseness

The ~25-line body is lean and dense: it assumes Claude's competence throughout ("Send one complete JSON request on standard input, close the input stream, and parse the single JSON response"), explains nothing Claude already knows, and every sentence carries operational content. Nothing reads as padding.

5 / 5

Actionability

Concrete guidance exists — the exact CLI invocation `node <skill-dir>/dist/550w-skill.cjs`, a worked example `{"action":"queryCredits","params":{}}`, env var names, action names, and param names like `params.filePath` — but the body's single executable target `dist/550w-skill.cjs` is absent from the bundle (no dist/, scripts/, or assets/ exist), so the primary instruction cannot actually run as shipped. No inline request example is given for the main workflow actions either; their shapes are deferred to the reference file. Not a 4 because the missing entry-point script is more than a minor gap.

3 / 5

Workflow Clarity

The sequence is clear: obtain credentials → disclose data upload before the first media operation → run the action → keep the task ID → poll via `taskDetail`/`taskList` → confirm before `deleteTask`. Checkpoints are mostly explicit: "A timeout does not prove submission failed: check the existing task before considering another billed submission", confirmation required before destructive delete, and per-unit billing stated before a batch. Not a 5: batch image processing ("process multiple images serially") lacks an explicit per-image verify step, so the feedback loop for batches is implied rather than spelled out.

4 / 5

Progressive Disclosure

A concise overview body with two well-signaled, purpose-stated references ([API contract](references/api-contract.md), [agent compatibility](references/agent-compatibility.md)) that both exist and are one level deep with no nested references. Not a 5: the body points to a bundled file `dist/550w-skill.cjs` that is not present in the bundle, leaving a dangling path alongside otherwise clean navigation.

4 / 5

Total

16

/

20

Passed

Description

70%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 tight, third-person, domain-specific description with concrete capabilities and strong natural trigger terms. Its main weakness is the absence of any "Use when..." trigger clause, which caps completeness and leaves activation guidance implicit.

Suggestions

Add an explicit trigger clause, e.g. "Use when the user wants to remove burned-in/hardcoded subtitles from a video, strip a watermark from a TikTok or short-video share link, clean a watermark or text from an image, or check 550W task status and credits."

Include common synonyms and formats users would say naturally: "burned-in subtitles", ".mp4/.mov", "PNG/JPG", "remove text from photo".

Make the auxiliary capability more concrete: replace "Also check tasks and credits" with "Query task status, task history, and remaining credits."

DimensionReasoningScore

Specificity

"Remove hardcoded subtitles from user-selected videos, watermarks from TikTok and other supported short-video share links, or watermarks and unwanted text from selected images" lists several concrete, distinct capabilities plus "check tasks and credits". Not a 5: the last clause is generic, no media formats or file extensions are named, and "check tasks and credits" leaves auxiliary capabilities under-specified.

4 / 5

Completeness

The "what" is clear and concrete (subtitle removal, share-link watermark removal, image watermark/text removal, task and credit checks), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric guidelines. The "when" is only weakly implied by the capability list itself.

3 / 5

Trigger Term Quality

Natural user phrases are present: "hardcoded subtitles", "watermarks", "TikTok", "short-video share links", "unwanted text", "videos", "images". Not a 5: misses common synonyms and extensions users would actually say — "burned-in subtitles", ".mp4", ".mov", "remove text from image/video", "clean up video".

4 / 5

Distinctiveness Conflict Risk

"hardcoded subtitles", "TikTok", "short-video share links", and "550W Open API" carve out a clear niche with distinct triggers and minimal overlap with other media skills. No generic phrasing like "helps with videos" that would collide with adjacent skills.

5 / 5

Total

16

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

sunshinehu/550w-ai-subtitle-remover

Path

claude-plugin/550w-ai-subtitle-remover/skills/ai-subtitle-remover/SKILL.md

Reviewed

Table of Contents

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.