메뉴
Pre-digest long inputs (video transcripts, CI logs, big files, web pages) with the LOCAL model (mlx_lm.server) so the Claude session reads a short digest + pointer instead of the full content — token saving on every long read.
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
Autonomous execution under hard caps — run ONE long objective hands-off, OR drive the planned effort/flow set in parallel. On an APPROVED plan the captain DECIDES, MERGES each wave, and CONTINUES; it pings the human ONLY on a genuine blocker. THIN wrapper over the built-in /loop + the kit's orchestrate scripts; never a new engine.
Close an effort in ONE op — snapshot each sub-issue's diff BEFORE squash (work record), capture/judge/sync effort metrics, squash-merge the effort PR, close the parent + ALL native sub-issues, set the board Status=Done for each, garbage-collect the worktree/branch, run the optional knowledge-ingest hook, and finish with the kit-sync drift check.
Start an effort — thin caller of `cckit effort start <N>`. Creates the `effort/<N>-<slug>` integration branch + isolated bootstrapped worktree from the base branch, then marks the parent issue In Progress on the board. Sub-issues later branch from this effort branch or commit directly on it.
Start work on a GitHub issue — thin caller of `cckit start <N>`. Creates an isolated, bootstrapped worktree + `<kind>/<N>-<slug>` branch from the base branch and marks the issue In Progress on the board (if enabled). Worktree-first — there is no in-place mode.
Create an effort — a parent GitHub issue using the effort-model template (Goal / Scope / For agents / Verification) plus N native GitHub sub-issues linked via the sub-issues REST API, all added to the project board. The parent issue IS the plan.
Open the ONE pull request for an effort — `effort/<N>` → base branch — with a rich human-facing review body plus a `## For agents` section listing the file paths/entry points. Title `[Role] [#N] <title>`. Inherits labels + milestone from the parent issue and moves it to In Review on the board.
| name | kit-digest |
| description | Pre-digest long inputs (video transcripts, CI logs, big files, web pages) with the LOCAL model (mlx_lm.server) so the Claude session reads a short digest + pointer instead of the full content — token saving on every long read. |
| when_to_use | Before reading any long input into the session — a YouTube/video transcript, a CI failure log, a file or page over ~2k words. Especially "procesa este video/transcript/log". If the local server is down, the script says so — fall back to reading the original directly. |
Plugin-direct skill — the script + local-model helper resolve from ${CLAUDE_PLUGIN_ROOT}.
Delegates to ${CLAUDE_PLUGIN_ROOT}/scripts/kit-digest.sh — acquires the text (file, URL, or
YouTube video via yt-dlp subtitles), chunks it (~2500 words), digests each chunk on the local
model (${CLAUDE_PLUGIN_ROOT}/scripts/lib/kit-local.sh → mlx_lm.server, $0 API), merges, and
prints a digest of at most ~1500 tokens plus a pointer to the original for selective deep-dives.
"${CLAUDE_PLUGIN_ROOT}/scripts/kit-digest.sh" <path|url> [--focus "<topic>"] [--lang es|en]
| Flag | Description |
|---|---|
--focus "<topic>" | Prioritize content related to a topic (e.g. --focus "claude-kit") |
--lang es|en | Digest language (default es) |
| Code | Meaning | What Claude does |
|---|---|---|
| 0 | Digest printed | Use the digest; read the original only for targeted deep-dives |
| 2 | Local server down | Tell the user (start command is in stderr) and read the original directly — current behavior, no digest |
| 1 | Input error (not found, no subs, empty) | Surface the error verbatim |
mlx_lm.server running (see the kit-local.sh header), yt-dlp only for YouTube URLs.