跳到正文
cccskills
免费在 GitHub 公开

bindry-sync

Compile skills from a Bindry Binder into a project’s Codex skills folder. Choose a pinned snapshot or live mode, and use a Binder ID, export URL, or saved export file as the source.

中文简介正在准备中。当前显示英语简介。

原文查看原文说明

Compile a bound Bindry Binder into Codex skills for this project, so its Skills load on demand instead of sitting in every prompt. Use when the user asks to sync, install, or update Bindry Binder skills for this project.

hashgraph-online/awesome-codex-plugins星标 1,422(GitHub 1,422 + cccskills 0)

实测评分 未实测

查看安装方法

适用场景

  • Install skills from a Bindry Binder
  • Refresh previously compiled skills
  • Compile a specific published Library version

包含的文件(1)

  • SKILL.md4.2 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Compile a Bindry Binder into Codex skills so its Skills load on demand instead of sitting in every prompt. Invoke explicitly with $bindry-sync, or let it trigger automatically when the user asks to sync/update Bindry skills.

  1. Determine the source from what the user actually said, in order:
    • A Binder id (GUID) or a full export URL they gave you — sync live from the Bindry API.
    • A file path they gave you — compile from that local export file.
    • Neither given: check for ./bindry.config.json in the current project (written by a previous sync) and reuse the Binder id it remembers.
    • None of those apply — ask the user for a Binder id (from Bindry → Binder → Export → target "codex") or a saved export file path, then stop.
  2. For a live sync, also need --api-base <url> (the Bindry API's base URL — ask the user if not obvious from context) and, for a private Binder, --token <api-key> (generated from the workspace's team page in Bindry, not Account settings). A published Library Binder (referenced by slug or GUID) needs no token at all — it resolves through the public Library's anonymous export route. Only your own private Binders require one.
  3. Default is --mode pinned (a static snapshot). If the user wants this Binder to always stay current instead — --mode live — first confirm the Bindry MCP server is connected (check the configured MCP servers for one named bindry). If it isn't connected, tell the user to use the bindry-connect skill first and stop; compiling live skills with no live connection would produce skills that always fail when used. --mode live also only works from a live Binder id/URL, not a local export file — its Skills need real GUID ids, which a hand-written or example export file won't have.
  4. If the user wants a specific published version of a Library Binder rather than its latest — e.g. rules for the framework release they are actually on — pass --version <v>. It is remembered in bindry.config.json, so later syncs stay on that version until --version latest removes the pin. A version pin only applies to a Library Binder (your own workspace Binder always compiles from its current composition), and an unknown version fails with a message naming the versions that do exist rather than quietly serving the latest.
  5. Locate this plugin's own installed root first (e.g. list installed Codex plugins and find the path for bindry@<marketplace>), then run: node <resolved-plugin-root>/scripts/compile-binder.mjs [<binder-id-or-file>] --out .agents/skills [--api-base <url>] [--token <api-key>] [--mode pinned|live] [--version <v>] If you can't locate scripts/compile-binder.mjs under this plugin's installed root, stop and say so plainly rather than guessing a path.
  6. Report exactly what the script printed: how many skills were written and where. If it failed with an authentication error, tell the user plainly what it said — don't guess or retry silently. If it wrote a bindry.config.json, mention that future syncs in this project can omit the Binder id (and remember the mode too).
  7. Remind the user to re-run this whenever a pinned Binder changes in Bindry — each pinned skill is tagged with a Skill version in an HTML comment, so the bindry-check skill can diff against it. A live skill never goes stale — it calls the Bindry MCP server for current content every time it's used — so there's nothing to re-sync for it, only re-run if the Binder's Skill list itself changed (added/removed Skills).
  8. If any Skill carries a file attachment (e.g. an exact logo to place), the script downloads it into that skill's assets/ folder alongside SKILL.md — in both pinned and live mode, since the live MCP tools don't serve binary files. If the script printed a warning that it skipped an asset, mention that plainly too — that skill still compiled, just without the file, and needs an --api-base/network fix before it will download.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Plan worthwhile local tasks by checking dependencies, effects, resources, and platform boundaries, then execute the exact approved plan. It can use safe parallel work or suitable Apple-silicon backends when the project exposes them.

中文简介正在准备中。当前显示英语简介。

  • Speed up independent local build or test steps
  • Plan a task with several dependency-linked operations
  • Choose an Apple-silicon backend for supported numerical or media work
hashgraph-online/awesome-codex-plugins14222026年10月27日 更新

add

免费

Record deferred engineering decisions and temporary workarounds in a debt registry, with context such as effort, ongoing cost, affected area, and payoff trigger. Entries receive short labels so they can be reviewed or dropped later.

中文简介正在准备中。当前显示英语简介。

  • Logging a TODO or unfinished implementation
  • Recording a temporary default or bypassed check
  • Tracking a swallowed error or loosened type
hashgraph-online/awesome-codex-plugins14222026年10月28日 更新

Generate or compare AI agent code for LangChain, EINO, AutoGen, AgentScope, and CrewAI using framework-specific templates, API mappings, conventions, and project structures. Reference files include verification dates to help identify potentially outdated guidance.

中文简介正在准备中。当前显示英语简介。

  • Creating an agent project for a chosen framework
  • Comparing APIs across supported agent frameworks
  • Referencing framework-specific tool and memory patterns
hashgraph-online/awesome-codex-plugins14222026年10月26日 更新

Investigate a reproducible bug, test failure, or unexpected page behavior, then make a narrowly scoped fix. The workflow calls for a regression test before changing code, verifies the fix, and reports remaining risks and evidence while avoiding unrelated refactoring or feature work.

中文简介正在准备中。当前显示英语简介。

  • Finding the cause of a failing test
  • Fixing an unexpected page or feature behavior
  • Adding a regression test before a minimal code change
hashgraph-online/awesome-codex-plugins14222026年10月28日 更新

Build an AI agent project through six collaborative stages, from requirements and architecture to tools, prompts, code generation, and verification. Start framework-independent, then generate a runnable project for frameworks such as LangChain, EINO, AutoGen, AgentScope, or CrewAI.

中文简介正在准备中。当前显示英语简介。

  • Designing an AI agent from a business need
  • Generating a runnable agent project
  • Planning tools, prompts, and orchestration
hashgraph-online/awesome-codex-plugins14222026年10月26日 更新

Audit a project or learning workflow for scope drift, conflicting plans, inconsistent names, and code quality concerns. The skill reads project files and code, then reports evidence, severity, and suggested next steps without changing files.

中文简介正在准备中。当前显示英语简介。

  • Checking whether project work has drifted from its agreed scope
  • Finding conflicting plans or inconsistent naming
  • Reviewing code against project decisions without editing it
hashgraph-online/awesome-codex-plugins14222026年10月28日 更新

查看 hashgraph-online 的全部技能

报告此技能的问题