Content
88%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is lean, highly actionable with exact FFmpeg chains and a working example, and clearly sequenced with validation and verification checkpoints; progressive disclosure is good with a real referenced script.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Efficient and assumes Claude's competence — no padding explaining FFmpeg or loudnorm — but the Purpose paragraph's narrative ('round 1 music too loud, round 2 VO too quiet...') is motivational context that could be trimmed; minor instances of over-explanation keep it just below 5. | 4 / 5 |
Actionability | Fully executable: exact filter_complex chain syntax ('loudnorm=I=...:TP=...:LRA=...,adelay=<start>|<start>,volume=<vo_volume>'), a real copy-paste example invocation, and a concrete reference to scripts/mix.sh (which exists) covering the common case. | 5 / 5 |
Workflow Clarity | Clear five-step sequence with explicit validation at step 1 ('--vo and --vo-starts must be same length') and verification at step 5 (ffprobe loudness probe, verification.md pass/fail), plus Quality Checks and Failure Modes covering error recovery — well above the destructive/batch cap of 3. | 5 / 5 |
Progressive Disclosure | Well-organized into labeled sections (Purpose, Inputs, Workflow, Output, Quality Checks, Failure Modes, Example) with a clearly signaled one-level reference to scripts/mix.sh used in the example; minor gaps only — most parameter detail is appropriately inline rather than split into separate files. | 4 / 5 |
Total | 18 / 20 Passed |