Overcast media bug triage

Skill kdr/overcast/skills/overcast-media-bug-triage

Analyze screen recordings, product demos, customer support videos, and audio notes into actionable, cited bug reports for coding agents.From its SKILL.md

Install

npx -y skills add kdr/overcast --skill overcast-media-bug-triage

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

12 stars12 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
runs commandsInstructs the agent to run 8 commands, including `overcast doctor --json` and 7 more.

SKILL.md

1.8 KB, 388 tokens by cl100k_base, as published. Nobody here has run it

overcast-media-bug-triage

Use this skill when media evidence should become a bug report, reproduction steps, or engineering triage notes. Use the broad overcast skill and overcast/reference/verbs.md for exact command flags.

Workflow

overcast doctor --json
overcast case init --json
overcast case setup --yes --json
overcast watch ./screen-recording.mp4 --json
overcast listen ./screen-recording.mp4 --describe --json
overcast see frame://<record-id>@<seconds> --ocr --json
overcast note "observed UI state or suspected failure" --ref <record-id> --at <time-range> --json
overcast ask "summarize the bug with reproduction steps and citations" --json
overcast brief --full --export ./bug-brief.md --json   # --full: this flow wants the verbatim evidence timeline (the default brief is short)

Use watch for screen recordings and demos. Add listen --describe when spoken narration, audio cues, or support-call context matters. Use see --ocr on key frames when UI text, error messages, button labels, or form values are important.

Output

Produce a cited bug summary with:

observed behavior with timestamps;
expected behavior when it is inferable from the media or product context;
reproduction steps grounded in record.id and media.at;
UI text or OCR evidence from see --ocr;
open questions when the media is ambiguous.

Evidence Rules

Keep observed media facts separate from engineering inference. Add human observations with note. Prefer ask and brief for synthesis; use case memory get to page large watch or listen fields when exact timeline text is needed.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most debug triage skills give in 388 tokens

Counted across 1,020 of the 1,639 authors here whose files we hold, read 2026-09-06

Find root cause before attempting any fixin 134 of 1020, across 118 files
Create a failing test case before implementing a fixin 109 of 1020, across 95 files
Read error messages and stack traces completelyin 102 of 1020, across 88 files
Reproduce the issue consistently before investigatingin 90 of 1020, across 77 files
Make the smallest possible change to test a hypothesisin 90 of 1020, across 76 files
Trace data flow backward to find the sourcein 84 of 1020, across 70 files
Form a single hypothesis before testingin 78 of 1020, across 64 files
Implement only one fix at a timein 76 of 1020, across 63 files
Question the architecture if three fixes failin 73 of 1020, across 59 files
Add diagnostic instrumentation at component boundariesin 68 of 1020, across 56 files
Compare broken code against working examplesin 68 of 1020, across 57 files
Write a regression test before applying the fixin 62 of 1020, across 55 files

Said here and by no other author read

initialize the case with overcast
watch screen recordings for visual evidence
listen to audio for narration context
use ocr on key frames for ui text
add observations with note command
synthesize findings with ask command

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.