CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

changelog-video

Turn a weekly changelog .md into a finished branded changelog video (square 1080, ~45-60s, Annie VO, animated brand background, mock-UI visualizations, lowkey captions). Use when the user provides a changelog/digest markdown and wants the weekly video, or says "changelog video". Self-contained — fonts, background, lexicon, and scripts ship in this skill.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

The canonical home for this skill is changelog-video in heygen-com/hyperframes

SKILL.md
Quality
Evals
Security

Quality

Content

81%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 unusually executable skill: exact commands, pinned assets and voice, hard validation gates with fallback paths, and well-integrated one-level references. The two real defects are the missing `examples/master-skeleton.html` (breaks step 0's foundational copy command) and repeated restatement of the same critical rules across sections and the anti-patterns table.

Suggestions

Ship `examples/master-skeleton.html` in the bundle (or point step 0 to a real path) — the scaffold copy is the foundation of steps 0 and 5 and currently fails as written.

Consolidate the scaffold/caption rules into one authoritative statement plus one anti-pattern row each; the same warning currently appears in step 0, the recovery paragraph, and two table rows.

Trim the parenthetical history (e.g. "This is exactly what went wrong on the Jul 13-20 v4 build") — the gate rule stands without the dated anecdote.

DimensionReasoningScore

Conciseness

The body is directive and assumes competence — no explaining what ffmpeg, TTS, or forced alignment are — and every section instructs rather than describes. Not a 5 because critical rules are restated several times in slightly different forms: the scaffold rule appears in the step 0 bold warning, the "If you catch yourself reaching for cp..." recovery paragraph, and two anti-pattern table rows; likewise captions are emphasized in steps 4, 5, 6 and two more table rows. These repetitions could be consolidated without losing the gate.

4 / 5

Actionability

Commands are copy-paste ready with exact flags and pinned values: the ffmpeg background encode ("-vf scale=1080:1080,fps=30,eq=saturation=0.72..."), the TTS call with pinned voice ID "330290724a1b470fb63153f34d4c0183", the uvx whisper fallback, and the full `check --caption-zone` parameter string. Not a 5 because step 0's `cp <SKILL_DIR>/examples/master-skeleton.html project/index.html` references a file that does not exist in the bundle (no `examples/` directory ships), so the foundational copy command fails as written.

4 / 5

Workflow Clarity

Steps 0-6 are clearly sequenced with explicit validation gates and feedback loops: "resolve every one before building" for MISMATCH warnings, the "Word-timings are a hard gate" check with a whisper fallback path, "all green before presenting" in step 6, and the hard-fail caption presence gate with a named recovery step ("re-check step 5's LINES population"). This matches the anchor of clear sequence plus explicit validation and error-recovery loops.

5 / 5

Progressive Disclosure

The body is a well-signaled router: `references/build-spec.md` ("read end-to-end"), `references/visualization-registry.md` ("route every theme/item through... BEFORE writing the script"), `references/script-voice.md`, `references/lexicon.json`, and `scripts/align-captions.mjs` all exist in the bundle, one level deep, each with a clear when-to-read cue. Not a 5 because the referenced `examples/master-skeleton.html` — the file step 0 and step 5 depend on entirely — is missing from the bundle, a dangling critical reference.

4 / 5

Total

17

/

20

Passed

Description

87%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 description: concrete deliverable specs, an explicit 'Use when' trigger clause including the literal user phrase 'changelog video', and a clearly distinct niche. It falls just short of top marks on action breadth (one transformation verb rather than multiple listed actions) and synonym coverage (no 'release notes' / 'weekly recap' variants).

Suggestions

Enumerate the constituent actions (parse the changelog, route visualizations, generate VO + captions, build and gate the composition) to lift specificity from one transformation verb to multiple concrete actions.

Add one or two natural synonym triggers such as 'release notes video' or 'weekly digest video' so users phrasing the request differently still hit the skill.

DimensionReasoningScore

Specificity

"Turn a weekly changelog .md into a finished branded changelog video" names the core action with highly concrete output specs ("square 1080, ~45-60s, Annie VO, animated brand background, mock-UI visualizations, lowkey captions"), plus the concrete fact that "fonts, background, lexicon, and scripts ship in this skill". Not a 5 because it describes one transformation with output properties rather than multiple distinct actions (parse, script, VO, build, verify are left implicit); not a 3 because the deliverable characteristics are far more concrete than "Names domain and 1-2 concrete actions".

4 / 5

Completeness

Explicitly answers both: what — "Turn a weekly changelog .md into a finished branded changelog video" with concrete output parameters — and when — "Use when the user provides a changelog/digest markdown and wants the weekly video, or says 'changelog video'". This matches the anchor for a clear 'what' plus explicit 'when' with concrete trigger phrases; a missing 'Use when' clause is not an issue here.

5 / 5

Trigger Term Quality

Natural triggers are strong: "changelog video" (the exact phrase a user would say), "changelog/digest markdown", "wants the weekly video". Not a 5 because common synonyms like "release notes video", "weekly recap", or "digest video" are missing; clearly above 3 since the literal quote trigger "says 'changelog video'" covers the primary phrasing users would use.

4 / 5

Distinctiveness Conflict Risk

"Changelog video" is a clear niche with distinct triggers — the combination of changelog markdown input, branded video output, named voice ("Annie VO"), and format ("square 1080, ~45-60s") is unlike any generic skill and would not fire for the wrong skill. Minimal conflict risk.

5 / 5

Total

18

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

referenced_paths_exist

Referenced path issues: 1 missing

Warning

Total

13

/

16

Passed

Repository

heygen-com/hyperframes

Path

.claude/skills/changelog-video/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.