CtrlK
BlogDocsLog inGet started
Tessl Logo

render-hook-replacement

Replace an existing video's opening with a supplied clip or free kinetic text hook while retaining and verifying every original body frame, audio, captions and ending. Use for surgical hook tests rather than regenerating the complete ad.

68

Quality

86%

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

SKILL.md
Quality
Evals
Security

Quality

Content

77%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 dense, highly disciplined instruction body with an exemplary validation-gated workflow and concrete numeric tolerances, but its progressive-disclosure structure is partly broken: config.example.json and tests/ are referenced yet absent, and the Related links are non-navigable markers. Minor internal redundancy (Human version, Quality Checks) keeps it from full token efficiency.

Suggestions

Add the actual config.example.json (or an inline example config) to the bundle so the Inputs section's format reference and the demo flow resolve for a fresh package.

Ship or remove the tests/ directory referenced by `python3 -m unittest discover -s tests -v` in the Free demonstration — as written, step 3 of the demo fails in a fresh checkout.

Convert the [[references::...]] / [[used-by::...]] markers in Related into real markdown links to existing files (or drop them), and fold the duplicated Quality Checks / Human-version content into the Workflow to remove redundancy.

DimensionReasoningScore

Conciseness

The body is lean and assumes competence: no explanation of what ffmpeg or PSNR are, commands given bare, tolerances stated as numbers ("PSNR ≥45dB", "RMS error ≤0.08", "correlation ≥0.99"). It is not a 5 because of real redundancy: the "Human version" preface restates the agent content, the Quality Checks section repeats verification rules already stated in Workflow steps 2–5, and the audio-tail rule appears in both Inputs and Failure Modes.

4 / 5

Actionability

Copy-paste commands for preflight, the main run, the demo, and the pack wrapper, plus a fully enumerated manifest schema and config fields (source.sha256, hook_end_sec, words, captions, output). It falls short of anchor 5 because the config format is never shown completely — the referenced [config.example.json](config.example.json) does not exist in the bundle, and the demo's `python3 -m unittest discover -s tests -v` references a `tests/` directory that is also absent, so the "free demonstration" cannot run as written from a fresh package folder.

4 / 5

Workflow Clarity

A five-step numbered workflow with explicit validation checkpoints: watch/confirm intake, preflight before any encoding, manifest booleans plus PSNR/RMS/correlation verification, watch the actual output, and a fix-and-re-run feedback loop ("Fix only the opening and re-run to a new path") backed by a Failure Modes table mapping each problem to a specific action. This matches anchor 5's clear sequence with explicit validation and error-recovery loops; the destructive/batch cap does not apply since verification gates are present.

5 / 5

Progressive Disclosure

The one real reference file (references/hook-replacement.md) is clearly signaled from Purpose and exists, and sections are well-organized. But disclosure is structurally undermined by unresolved paths: config.example.json and the tests/ directory referenced in the body are missing from the bundle, and the "Related" section uses opaque wiki-style markers ([[references::preflight-concat]], [[used-by::...]]) that resolve to nothing a reader can open. This sits between anchor 4 (mostly clear references, minor gaps) and anchor 3 — closer to 3 because two inline references and the entire Related section do not navigate anywhere.

3 / 5

Total

16

/

20

Passed

Description

92%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, niche-specific, third-person, with an explicit "Use for..." trigger clause and a useful contrast against full-ad regeneration. Only minor gap is missing common synonyms ("intro", ".mp4") in the trigger vocabulary.

DimensionReasoningScore

Specificity

"Replace an existing video's opening with a supplied clip or free kinetic text hook while retaining and verifying every original body frame, audio, captions and ending" lists multiple concrete actions (replace opening, supply clip, kinetic text hook, retain body/audio/captions/ending, verify) with comprehensive coverage of the skill's scope. It exceeds anchor 4 ("several specific actions; minor gaps") because nothing meaningful in the capability is left unstated; the compressed jargon ("free kinetic text hook") is a phrasing choice, not a coverage gap.

5 / 5

Completeness

The "what" is explicit (replace opening with clip or kinetic text while retaining/verifying body, audio, captions, ending) and the "when" is an explicit trigger clause: "Use for surgical hook tests rather than regenerating the complete ad" — which also carries a contrast that sharpens when it applies. This clearly matches anchor 5 (both what and when with concrete trigger phrases) rather than anchor 4, where the "when" would be only implicit or less explicit.

5 / 5

Trigger Term Quality

Natural terms present: "video's opening", "supplied clip", "hook", "captions", "ending", "surgical hook tests" — good keyword coverage matching what a user would say ("replace the opening/hook of this video"). It falls short of anchor 5 because common synonyms and file extensions are missing ("intro", "first few seconds", ".mp4", "ad creative"), so a few natural trigger phrasings would not match.

4 / 5

Distinctiveness Conflict Risk

A clear niche — surgical video opening/hook replacement with body preservation verification — with distinct triggers (hook replacement, opening swap, retained body/captions) that would rarely collide with general editing or generation skills. The "rather than regenerating the complete ad" clause actively distinguishes it from whole-video regeneration skills, matching anchor 5.

5 / 5

Total

19

/

20

Passed

Validation

87%

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

Validation — 14 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

relative_links

Relative link issues: 1 missing

Warning

Total

14

/

16

Passed

Repository
gooseworks-ai/goose-skills
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.