원클릭으로
screen-record
Capture screenshots or recordings from windows or displays through the nils-cli `screen-record` command.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Capture screenshots or recordings from windows or displays through the nils-cli `screen-record` command.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Operate and test a local or SSH-reachable macOS desktop through nils-cli macos-agent, with AX-first actions, screenshots, scenarios, explicit postconditions, guarded replay, and privacy-preserving journals.
Review a code change through an internally selected quick, focused, specialist, follow-up, or pre-merge mode and return evidence-grounded findings.
Open or resume one shared dispatch plan issue, coordinate independently reviewed lane PRs, integrate approved work, and close through strict gates.
Open or resume one lightweight issue-backed plan tracker and carry it through implementation, review, PR delivery, strict closeout, and archive handoff.
Deliver GitHub pull requests or GitLab merge requests end to end through the released nils-cli `forge-cli pr deliver` macro.
Review a code change through an internally selected quick, focused, specialist, follow-up, or pre-merge mode and return evidence-grounded findings.
| name | screen-record |
| description | Capture screenshots or recordings from windows or displays through the nils-cli `screen-record` command. |
Prereqs:
screen-record is installed from the released nils-cli package and available on PATH.Inputs:
Outputs:
Failure modes:
The user requests a screenshot or recording. The workflow selects preflight, target discovery, capture, and diagnostics from the requested claim. When no path is supplied, allocate one project-scoped output directory; evidence metadata and diagnostics are internal bookkeeping and are reported only when they support the result or explain a failure.
Use the released CLI directly:
screen-record --preflight
screen-record --list-windows
RUN_DIR="$(agent-out project --topic screen-record --repo "$PWD" --mkdir)"
screen-record --screenshot --active-window --path "$RUN_DIR/window.png"
screen-record --display --duration 10 --path "$RUN_DIR/session.mov" --metadata-out "$RUN_DIR/session.json"
--preflight when host permissions or capture backend status is uncertain.--list-windows, --list-displays, or --list-apps before selecting a target by id or name.--path for evidence artifacts that need to be linked from validation.--diagnostics-out for failed or flaky recording checks.screen-record owns capture and metadata mechanics. The caller owns choosing a non-sensitive target and deciding whether artifacts are temporary evidence or durable records.