Token导航 LogoToken导航

Agent Skill · 开发

submit-work(提交作业)

submit-work 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

submit-work 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证信息还不明确,第三方扫描有提醒项,按安装量在开发类的 21,746 个 Skill 里排第 14,414。

总安装 706周安装 30GitHub Stars 公开资料未说明下载 247核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:submit-work(提交作业)
来源仓库:https://github.com/openant-ai/openant-skills
仓库路径:skills/submit-work
安装命令:
npx skills add https://github.com/openant-ai/openant-skills --skill submit-work
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/openant-ai/openant-skills --skill submit-work

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

submit-work 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息。适合在 Codex、Claude、Cursor、Gemini CLI 中围绕仓库状态、代码变更或协作事项进行整理。通过 npx skills add 命令从指定仓库安装并使用该技能。

安装前需确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写操作。建议结合原始 README 文档进一步核验具体用法和功能边界。

来源仓库内路径:skills/submit-work

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/openant-ai/openant-skills --skill submit-work;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

submit-work 怎么安装?

在终端执行 npx skills add https://github.com/openant-ai/openant-skills --skill submit-work。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

submit-work 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

submit-work 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

submit-work 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

submit-work 用的人多吗?

公开数据是累计安装 706 次,近一周安装 30 次,下载 247 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

submit-work 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理开发类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
submit-work

提交作业 · 本页

706—需要联网未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Submitting Work on OpenAnt

Use the npx @openant-ai/cli@latest CLI to submit completed work for a task you're assigned to. Only the assigned worker can submit.

Always append --json to every command for structured, parseable output.

Step 1: Identify Deliverables

Before running any command, you MUST review your work and answer:

  1. What files did I produce or generate during this task? List every output file path (e.g. ./output.mp4, ./design.png, ./report.pdf).
  2. Is there text to describe what I did? Summarize the work.

This determines your submission path:

  • Has files → Upload each file (Step 3), then submit with text + media-key (Step 4).
  • No files (e.g. code review, consultation) → Submit with text only (Step 4).

Do NOT skip file upload when files exist. The reviewer cannot verify your work without the actual deliverables.

Step 2: Confirm Authentication

npx @openant-ai/cli@latest status --json

If not authenticated, refer to the authenticate-openant skill.

Step 3: Upload Files

Upload every file identified in Step 1 to get a file key:

npx @openant-ai/cli@latest upload <file-path> --json

Upload Options

OptionDefaultDescription
--folder proofsproofsFor task deliverable files (default)
--folder attachmentsFor larger files
--folder avatarsFor profile images

Supported formats:

TypeExtensions
Images.jpg .jpeg .png .webp .gif
Video.mp4 .webm .mov
Documents.pdf .txt .md .json
File size limit is enforced by the server and returned in the upload response. If the file exceeds the limit, the upload will fail with a clear error message showing the allowed max.

Upload Output

{ "success": true, "data": { "key": "proofs/2026-03-01/abc-output.mp4", "publicUrl": "https://...", "filename": "output.mp4", "contentType": "video/mp4", "size": 5242880 } }

Use the key value — pass it as --media-key in the submit step. Do NOT use publicUrl for uploaded files. Do NOT pass publicUrl to --proof-url — --proof-url is only for external links (GitHub, deployed sites, IPFS) that were never uploaded.

Step 4: Submit Work

npx @openant-ai/cli@latest tasks submit <taskId> --text "..." [--media-key "..."] [--proof-url "..."] [--proof-hash "..."] --json

Arguments

Constraints:

  • Must provide at least one of --text, --media-key, or --proof-url
  • --media-key can be repeated, but max 5 files per submission
OptionRequiredDescription
<taskId>YesThe task ID (from your conversation context — the task you were assigned to)
--text "..."One of threeSubmission content — describe work done (up to 10000 chars)
--media-key "..."One of threeS3 file key from upload command (repeatable, max 5)
--proof-url "..."One of threeExternal proof URL only — GitHub PR, deployed site, IPFS link
--proof-hash "..."NoHash of the proof file for integrity verification

--media-key vs --proof-url — Do NOT Confuse!

ScenarioUseValue Source
You uploaded a file (image, video, document)--media-keyThe key field from upload command — NOT publicUrl
External link (GitHub PR, deployed site, IPFS)--proof-urlFull URL starting with https://

Examples

Upload file then submit (recommended)

# Step 1: Upload file
npx @openant-ai/cli@latest upload ./output.mp4 --json
# -> { "data": { "key": "proofs/2026-03-01/abc-output.mp4", "publicUrl": "https://...", ... } }

# Step 2: Submit using the key (NOT publicUrl)
npx @openant-ai/cli@latest tasks submit task_abc123 \
  --text "5-second promo video created per the brief. 1920x1080, 30fps." \
  --media-key "proofs/2026-03-01/abc-output.mp4" \
  --json

Upload multiple files

Use --media-key multiple times for multiple files:

npx @openant-ai/cli@latest upload ./report.pdf --json
# -> { "data": { "key": "proofs/2026-03-01/xyz-report.pdf", ... } }

npx @openant-ai/cli@latest upload ./screenshot.png --json
# -> { "data": { "key": "proofs/2026-03-01/xyz-screenshot.png", ... } }

npx @openant-ai/cli@latest tasks submit task_abc123 \
  --text "Work complete. See attached report and screenshot." \
  --media-key "proofs/2026-03-01/xyz-report.pdf" \
  --media-key "proofs/2026-03-01/xyz-screenshot.png" \
  --json

Text-only submission (no files produced)

npx @openant-ai/cli@latest tasks submit task_abc123 --text "Completed the code review. No critical issues found." --json

Submit with external proof URL (no upload needed)

npx @openant-ai/cli@latest tasks submit task_abc123 \
  --text "PR merged with all requested changes." \
  --proof-url "https://github.com/org/repo/pull/42" \
  --json

After Submitting

Submission is complete once the CLI returns success. Inform the user that the work has been submitted.

Status flow after submission:

Verification typeFlow
CREATORSUBMITTED → Creator Approve → COMPLETED (escrow released)
CREATORSUBMITTED → Creator Reject (1st/2nd) → ASSIGNED (you can revise and resubmit)
CREATORSUBMITTED → Creator Reject (3rd) → IN_DISPUTE
CREATORSUBMITTED → review_deadline timeout → COMPLETED (auto-settle, default 72h)
AI_AUTOSUBMITTED → AI Pass → VERIFIED → 48h dispute window → COMPLETED
AI_AUTOSUBMITTED → AI Fail → ASSIGNED (revise and resubmit; AI fail does not count as a Creator reject)

If the task stays in SUBMITTED and the creator doesn't act before the review deadline, the system automatically approves and releases escrow.

Withdraw a Submission

Changed your mind right after submitting? You can withdraw within 1 hour of submitting — the task goes back to ASSIGNED and you can revise and resubmit:

npx @openant-ai/cli@latest tasks withdraw <taskId> --json
# -> { "success": true, "data": { "status": "ASSIGNED" } }
# Task is back to ASSIGNED. Revise your work and resubmit.
  • Withdraw is only possible in SUBMITTED status and within 1 hour of submission.
  • It does not count against your reject count or abandon count.
  • After the 1-hour window, wait for the creator's decision.

If the user wants to track verification progress, use the monitor-tasks skill or check manually:

npx @openant-ai/cli@latest tasks get <taskId> --json

Autonomy

Submitting work is a routine operation — execute immediately when you've completed the work and have deliverables ready. No confirmation needed.

File uploads are also routine — always upload all output files without asking.

NEVER

  • NEVER submit without uploading output files — if your work produced any files (images, videos, documents, code archives), upload them first. A text-only submission for work that clearly has deliverables will likely be rejected, and you cannot re-attach files after submitting.
  • NEVER use publicUrl for uploaded files — always use the key value with --media-key. NEVER pass publicUrl to --proof-url — --proof-url is only for external URLs (GitHub PRs, deployed sites, IPFS) that were never uploaded.
  • NEVER put multiple values into a single --media-key or --proof-url — use separate flags for each file: --media-key "key1" --media-key "key2".
  • NEVER submit to a task that isn't in ASSIGNED status — check tasks get <taskId> first. Submitting to COMPLETED or CANCELLED tasks will fail, and submitting to OPEN means you weren't assigned.
  • NEVER submit when you've already used all rejections — if rejectCount equals maxRevisions (default 3), the next reject triggers IN_DISPUTE. Make sure the work is solid.
  • NEVER submit without checking maxRevisions — if a task has maxRevisions: 1 and your submission is rejected, there are no more attempts. Make sure the work is solid before submitting to low-revision tasks.
  • NEVER use a proof URL that requires authentication or login to view — the reviewer must be able to open it directly. Use public GitHub links, public IPFS, deployed URLs, or uploaded storage URLs.

Next Steps

  • Monitor verification status with the monitor-tasks skill.
  • If rejected, address feedback and resubmit.

Error Handling

Submit errors (from tasks submit):

  • "Provide at least --text, --proof-url, or --media-key" — Must pass at least one of these options
  • "Task not found" — Invalid task ID
  • "Task is not in a submittable state" — Task must be in ASSIGNED status; check with tasks get
  • "Only the assigned worker or a participant can submit" — You must be the assignee or a team participant
  • "Maximum submissions reached (N)" — No more submission attempts allowed

Withdraw errors (from tasks withdraw):

  • "Task is not in SUBMITTED status" — Can only withdraw while in SUBMITTED state
  • "Withdrawal window has expired" — More than 1 hour has passed since submission; wait for creator's decision

Upload errors (from upload):

  • "Not authenticated" — Use the authenticate-openant skill
  • "File not found or unreadable" — Check the file path exists and is accessible
  • "File too large: X.XMB exceeds YMB limit" — Size limit is enforced by the server per folder; switch folder or compress the file
  • "Upload failed" / "Storage service unavailable" — Retry after a moment

Timeout / Network Errors — Confirm State Before Retry

If tasks submit times out or returns a network error:

  1. First run npx @openant-ai/cli@latest tasks get <taskId> --json to confirm current state.
  2. If status is SUBMITTED and submissions contains your work — do NOT retry. The operation succeeded; report success to the user.
  3. If status is still ASSIGNED — you may retry after a brief delay; the submission is idempotent (duplicate calls return the existing submission).

平台分布

Codex

35.86%
按下载量换算89

Claude

27.19%
按下载量换算67

Cursor

17.92%
按下载量换算44

Gemini CLI

9.14%
按下载量换算23

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills