CtrlK
BlogDocsLog inGet started
Tessl Logo

watching-issue-video-recordings

Use when a GitHub issue, PR, or bug report links a screen-recording video (github.com/user-attachments/assets/... .mov/.mp4/.webm) and you need to know what it shows. You cannot view video in context — this extracts still frames to a temp dir so you can Read them as images. Trigger whenever a task references a video/screen recording/attachment demonstrating a bug, or when text says "see the recording"/"linked video".

71

Quality

86%

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

SKILL.md
Quality
Evals
Security

Quality

Content

88%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.

A tight, highly actionable skill body with executable commands, a validated multi-step workflow, and clean sectioning. The only slack is a somewhat redundant Real-World Impact anecdote and minor overlap between Common Mistakes and the workflow steps.

DimensionReasoningScore

Conciseness

The body is efficient and assumes Claude's competence — it never explains what ffmpeg, fps, or frames are — with concrete commands throughout. Minor trimmable padding exists: the 'Real-World Impact' anecdote (~120 words) and some overlap between 'Common Mistakes' and the workflow steps, which keeps it just below anchor 5. It is not a 3 because there is no concept-explanation fluff and most tokens earn their place.

4 / 5

Actionability

Fully executable guidance: the exact invocation 'extract-frames.sh <url-or-path> [fps] [start] [duration]' plus a Quick Reference covering first-pass, zoom ('extract-frames.sh <url> 10 2 1.5'), local file, and cleanup ('rm -rf <WORKSPACE>', 'git status --porcelain') — matching anchor 5's copy-paste-ready coverage of common cases. It is not a 4 because no key execution detail is missing.

5 / 5

Workflow Clarity

A 6-step numbered checklist with elaboration, explicit validation (step 5 reconcile against the written repro, step 6 'git status --porcelain' repo-cleanliness check), and a feedback loop (step 4 zoom on ambiguity). Because validation IS present, the destructive/batch cap at 3 does not apply. It is not a 4 because checkpoints and an error-recovery loop are explicit, matching anchor 5.

5 / 5

Progressive Disclosure

Well-organized into clear sections (Overview, Workflow, Hard constraints, Quick Reference, Common Mistakes, Real-World Impact) with a single, well-signaled one-level-deep reference to the bundled 'scripts/extract-frames.sh' (verified present). It is not a 5 because the self-contained 'Real-World Impact' anecdote is slightly off-purpose and could be trimmed or moved, leaving a minor organization gap at anchor 4.

4 / 5

Total

18

/

20

Passed

Description

85%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 strong, highly targeted description with comprehensive trigger terms and a clear what/when structure. The only weakness is its second-person voice, which the rubric penalizes on specificity.

Suggestions

Rewrite in third person to avoid the specificity penalty, e.g. 'Extracts still frames from a screen-recording video linked in a GitHub issue/PR/bug report into a temp dir, so frames can be Read as images. Use when...' instead of 'you need to know' / 'you can Read them'.

Add one more concrete action (e.g. 'zoom-extracts a time window at higher fps' or 'cleans up the temp workspace') to lift specificity toward a 4 once the voice penalty is removed.

DimensionReasoningScore

Specificity

The description names the domain (issue/PR screen recordings) and one concrete action — 'extracts still frames to a temp dir so you can Read them as images' — matching the anchor-3 pattern of domain + 1-2 actions, but it is reduced by 1 per the rubric's second-person penalty ('you need to know', 'you cannot view video', 'you can Read them'). It is not a 1 because a concrete action is named, and not a 3 because the second-person voice triggers the explicit penalty.

2 / 5

Completeness

Explicitly answers both 'what' ('extracts still frames to a temp dir so you can Read them as images') and 'when' ('Use when a GitHub issue, PR, or bug report links a screen-recording video...', 'Trigger whenever...') with concrete trigger phrases — matching anchor 5. It is not a 4 because both halves are concrete and explicit rather than weakly implied.

5 / 5

Trigger Term Quality

Comprehensive natural-term coverage including synonyms and file extensions: 'video', 'screen recording', 'attachment', 'see the recording', 'linked video', and '.mov/.mp4/.webm' — matching the anchor-5 example. It is not below 5 because both everyday phrasings and extensions are present.

5 / 5

Distinctiveness Conflict Risk

Clear niche — screen recordings attached to GitHub issues/PRs/bug reports, with a distinct URL pattern (github.com/user-attachments/assets/...) and minimal overlap with other skills — matching anchor 5. It is not a 4 because the trigger is tightly scoped with no meaningful overlap risk.

5 / 5

Total

17

/

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
HHS/OPRE-OPS
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.