CtrlK
BlogDocsLog inGet started
Tessl Logo

andrebrov/broll-sourcing

Drop in one talking-head video and get it finished — transcribe it, search the web for supporting b-roll (article screenshots, memes, images, video clips), present the asset list for approval, then hand off to reel-builder to normalize audio, burn subtitles, cut the b-roll in, and export for Instagram and LinkedIn. Use when the user drops a video and wants illustrations/screenshots/memes added, asks to "edit this like CapCut", wants b-roll found for them, or wants a talking-head take prepped for Instagram/LinkedIn. Sourcing and approval only — all rendering belongs to the reel-builder skill.

74

Quality

93%

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

Overview
Quality
Evals
Security
Files

Quality

Content

85%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.

The body is an exceptionally actionable, well-sequenced pipeline with strong validation feedback loops and real bundle scripts. Its main weakness is progressive disclosure: a 460-line monolith that inlines several advanced topics instead of pointing to reference files.

Suggestions

Extract the Descript API section and the Transitions/Karaoke-caption details into reference files (e.g. references/descript.md, references/transitions.md) and link to them from SKILL.md so the main file reads as an overview.

Trim illustrative anecdotes (e.g. the 'three of twelve boundaries cut mid-word' and observed ASR mishear lists) to the underlying rule, keeping the load-bearing command/assertion but dropping the narrative.

Consider moving the 'What this pipeline cannot do locally' capability table into a reference doc, keeping only a one-line pointer in the main flow.

DimensionReasoningScore

Conciseness

The body is dense and high-signal, assuming Claude's competence (no basic explanations of ffmpeg/whisper), but at ~460 lines it carries a few illustrative anecdotes ('In one take, three of twelve boundaries cut mid-word...') and narrative color that could be trimmed without losing the rule.

4 / 5

Actionability

Copy-paste-ready bash, python, and ffmpeg commands appear throughout (init, mlx_whisper, plan_broll.py, magick pad, loudnorm, reframe_offset.py, export_variants.py), plus a concrete JSON slot spec — fully executable guidance covering the common cases.

5 / 5

Workflow Clarity

A clear Step 1–6 sequence with a top-of-file flow diagram and explicit validation feedback loops — 'Gate on the dependency check before scaffolding', 'verify_cut.py after any audio-driven edit — and believe it', the loudness gate 'reject anything outside -16..-13' — with fix-and-retry guidance for a destructive/batch render pipeline.

5 / 5

Progressive Disclosure

Structure and sectioning are clear and the referenced bundle scripts (transitions.py, reframe_offset.py, finish_reel.py, cut_boundaries.py) all exist in scripts/, but the ~460-line SKILL.md inlines advanced topics (Descript API, transitions styles, karaoke captions) that would more naturally live in one-level-deep reference files rather than the overview.

3 / 5

Total

17

/

20

Passed

Description

100%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.

The description is a strong, third-person statement that names concrete actions, gives explicit natural trigger phrases, and draws a clear boundary against a sibling skill. It satisfies the 'what', 'when', and 'where it stops' questions without padding.

DimensionReasoningScore

Specificity

Lists multiple concrete actions — 'transcribe it, search the web for supporting b-roll (article screenshots, memes, images, video clips), present the asset list for approval, then hand off to reel-builder to normalize audio, burn subtitles, cut the b-roll in, and export' — covering the full sourcing-to-handoff scope comprehensively.

5 / 5

Completeness

Explicitly answers both 'what' (transcribe, search, approve, hand off) and 'when' via a concrete 'Use when...' clause with multiple trigger phrases, plus a boundary statement ('Sourcing and approval only — all rendering belongs to the reel-builder skill').

5 / 5

Trigger Term Quality

Natural user phrasing is covered with synonyms and platform names — 'wants illustrations/screenshots/memes added', 'asks to "edit this like CapCut"', 'wants b-roll found for them', 'talking-head take prepped for Instagram/LinkedIn' — matching how a user would actually request this.

5 / 5

Distinctiveness Conflict Risk

Clear niche (b-roll sourcing + approval) with an explicit boundary against the sibling reel-builder skill and distinct triggers (CapCut, b-roll, memes, talking-head prep), minimizing overlap risk.

5 / 5

Total

20

/

20

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

15

/

16

Passed

Reviewed

Table of Contents