AGENT SKILL

harvest-playbook

Harvest a finished AKARI Video project's workflow into planning/playbook.md and planning/playbook.json, grouping media by role and leaving external media as empty placeholders. Use when asked to save a project's way of working as a reusable playbook.(旧名: record-playbook)

License 许可
MIT
最近更新
2026-10-22
Stars
★ 214
Forks
25

01 前言 Front Matter

name
harvest-playbook
description
Harvest a finished AKARI Video project's workflow into planning/playbook.md and planning/playbook.json, grouping media by role and leaving external media as empty placeholders. Use when asked to save a project's way of working as a reusable playbook.(旧名: record-playbook)

02 资源文件 2

03 SKILL.md 正文

Record a playbook

Language: Respond in the user's language — conversation, questions, approval prompts, and the reports and records you write follow the language the user writes in.

Hard rules

  1. A playbook is a document and inventory, not an automatic execution engine. Follow docs/contract-2026-10-15-playbook-v0.md.
  2. Read edit.json through a script, following docs/guides/edit-json-access.md. Do not read the entire file into agent context. Never copy source media, source paths, or the original edit into the playbook.
  3. Group sources by their role. Sum the duration hints for sources serving the same role. For own and site material, write an empty placeholder without a sample. Do not put own or site items in parts.
  4. Ask the person once for name, category, and license only. Use CC0-1.0 as the proposed license. Do not ask for each step, role, setting, or part. Fill uncertain wording with 要確認 and refine it from observed evidence.
  5. Keep the Markdown headings, including ## 内訳 after ## 必要な素材(仮枠), and the JSON schema in sync. For an existing playbook, increment playbookVersion. Use atomic writes.

Procedure

  1. Inspect decisions.json, the documents in planning/, and report types in .akari/reports/ for the steps and human checkpoints. If the steps cannot be recovered, use four defaults: place media, confirm script, refine cut, export. Mark uncertain what text 要確認.
  2. Extract aspect and fps from edit.json, and target duration from .akari/intake.json (target.duration_s), falling back to the used timeline length when no target is set. Caption style/position, default transition, and audio roles currently have no project-level default consumed when new work is created. Preserve those planning values from an existing planning/playbook.json when recording an applied playbook; otherwise infer transition from used cuts and audio roles from item gains where possible. Record settings as values to start the next project, not as a copy of every cut.
  3. Examine edit.json sources and available analysis.json tags or speaker observations to name placeholder roles. Use .akari/asset-references.json and .akari/style-usage.json to identify reusable parts and licenses. Keep only builtin, lab, and ai parts. Check used skills and recipe requires for required tools and skills. Derive beats from plan lines in planning/script.json when present; otherwise use chapter headings in captions or planning/plan.md together with visual cut boundaries. Write the ## 内訳 table in playbook.md.
  4. Run node skills/harvest-playbook/bin/record-playbook.mjs <project-root> --name <name> --category <category> [--license <license>] from a repository checkout, or use .claude/skills/harvest-playbook/bin/record-playbook.mjs from a scaffolded project. The helper writes deterministic structure and 要確認 defaults. If a video exists in exports/, it uses ffmpeg and ffprobe from PATH to extract 3–5 representative PNG frames into planning/playbook-preview/frames/ and records relative paths in preview.frames. AKARI_FFMPEG and AKARI_FFPROBE can override those commands. If no export or tool is available, or extraction fails, the helper warns and the outcome section says to record again after export. It clears old PNG frames before each extraction. You may review and replace frame choices while keeping 3–5 frames, the playbook aspect ratio, and a maximum long edge of 640 px. Refine the step wording and role names in both files where evidence supports better names; preserve role grouping and validate after editing with node packages/schemas/validate-playbook.mjs <project-root>/planning/playbook.json when the repository checkout is available.
  5. Show the playbook and preview frames to the person, then run node skills/harvest-playbook/bin/record-playbook.mjs <project-root> --install if they want the reviewed files copied to their local shelf at ~/.akari/playbooks/<id>/. This standalone install does not record again, so reviewed frame choices survive. The helper changes preview paths to preview/frames/0N.png in the shelf copy and replaces an existing copy only when playbookVersion is higher.
  6. Summarize the flow, placeholder count, part count, and required tools on one screen.

04 同仓库更多 Skills

AkariLabs
AkariLabs/akari-video

verify-project

Verify an AKARI Video project with the L0 / L1 / L2 verification ladder required by an AKARI Video task contract on the current Theia stack. Read when acceptance criteria specify levels such as L0 or L0+L1, to determine exactly what to run and how at each level.(旧名: verify)

★ 214
AkariLabs
AkariLabs/akari-video

verify-edit

Verify edit.json and optional analysis.json / captions.json / media with a deterministic CLI, then finish QA with frame inspection and reporting after PASS. Use immediately after writing or modifying edit.json, before export, and after applying review feedback.(旧名: edit-lint)

★ 214
AkariLabs
AkariLabs/akari-video

setup-remote

Set up remote access to approval reports and previews from another device such as a phone, and send footage to the workspace. Use doctor to assess Tailscale, guide installation and login (human actions), expose the preview server (default 4567) and approval report helper over tailnet-only HTTPS with tailscale serve, connect Taildrop reception to workspace inbox/, and verify access from another device. Use for requests to view reports on a phone, approve remotely, send footage from a phone, or set up remote access. Public internet exposure (funnel) is outside the default scope.

★ 214
AkariLabs
AkariLabs/akari-video

setup-project

Set up an AKARI Video project headlessly. Recursively copy `templates/project-default/`, record the template version, initialize git only when safe, and generate an HTML creation report. No app launch is needed. Use to create a video project or fill in an existing folder as an AKARI Video project.(旧名: create-project)

★ 214
AkariLabs
AkariLabs/akari-video

setup-library

Use for first-time AKARI Video setup or when the current project needs more usable assets. This first-run skill checks ffmpeg / whisper-cli (and can install with approval), reads the bundled catalog with akari assets to propose a starter pack, obtains explicit human approval, then acquires, places, validates, and updates INDEX entries. It also covers the audio library (BGM and sound effects): bulk acquisition of first-party AKARI Sounds from GitHub Releases, a free-source candidate HTML list with drop-folder matching or assisted acquisition for missing families, and keep/drop decisions in the listening gallery. Use for requests such as "I want more BGM" or "add sound effects to the library".(旧名: setup-audio-library)

★ 214
AkariLabs
AkariLabs/akari-video

setup-connections

Set up and inspect AKARI Video connections: AI-service keys, which models are allowed, cost-approval policy, the tool and media-service state, and 3D settings. Use for first-time setup, adding a provider or key later, checking what is usable (「何が使える?」 'what can I use?'), setting a key such as fal (「fal の鍵を設定したい」), 3D settings (「3D の設定を使いたい」), connection checks, and gates before paid generation or external publication. Model selection belongs to the production skills (they read the ledger and preferences); this skill decides allowed models and the cost policy. Works with `.akari/connections.json` and a free, read-only doctor.(旧名: manage-connections)(旧名: connect)

★ 214