MCP Server Builder
anthropics/skills
Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.
Environment-specific adaptation rules for EdgeOne Makers Skills running in sandboxed or restricted AI coding environments (e.g.
$ npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install TencentEdgeOne/edgeone-makers-tools edgeone-makers-env-adaption --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/TencentEdgeOne/edgeone-makers-tools.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/edgeone-makers-tools/references/makers-env-adaption .claude/skills/edgeone-makers-env-adaption && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "edgeone-makers-env-adaption" agent skill from https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaption into .claude/skills/edgeone-makers-env-adaption/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "edgeone-makers-env-adaption", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaptionType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install TencentEdgeOne/edgeone-makers-tools edgeone-makers-env-adaption --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentEdgeOne/edgeone-makers-tools.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/edgeone-makers-tools/references/makers-env-adaption .agents/skills/edgeone-makers-env-adaption && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "edgeone-makers-env-adaption" agent skill from https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaption into .agents/skills/edgeone-makers-env-adaption/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "edgeone-makers-env-adaption", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install TencentEdgeOne/edgeone-makers-tools edgeone-makers-env-adaption --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentEdgeOne/edgeone-makers-tools.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/edgeone-makers-tools/references/makers-env-adaption .cursor/skills/edgeone-makers-env-adaption && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "edgeone-makers-env-adaption" agent skill from https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaption into .cursor/skills/edgeone-makers-env-adaption/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "edgeone-makers-env-adaption", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/TencentEdgeOne/edgeone-makers-tools.git --path skills/edgeone-makers-tools/references/makers-env-adaption--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install TencentEdgeOne/edgeone-makers-tools edgeone-makers-env-adaption --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentEdgeOne/edgeone-makers-tools.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/edgeone-makers-tools/references/makers-env-adaption .gemini/skills/edgeone-makers-env-adaption && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "edgeone-makers-env-adaption" agent skill from https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaption into .gemini/skills/edgeone-makers-env-adaption/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "edgeone-makers-env-adaption", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install TencentEdgeOne/edgeone-makers-tools edgeone-makers-env-adaptionInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/TencentEdgeOne/edgeone-makers-tools.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/edgeone-makers-tools/references/makers-env-adaption .github/skills/edgeone-makers-env-adaption && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "edgeone-makers-env-adaption" agent skill from https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaption into .github/skills/edgeone-makers-env-adaption/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "edgeone-makers-env-adaption", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install TencentEdgeOne/edgeone-makers-tools edgeone-makers-env-adaption --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentEdgeOne/edgeone-makers-tools.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/edgeone-makers-tools/references/makers-env-adaption .opencode/skills/edgeone-makers-env-adaption && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "edgeone-makers-env-adaption" agent skill from https://github.com/TencentEdgeOne/edgeone-makers-tools/tree/main/skills/edgeone-makers-tools/references/makers-env-adaption into .opencode/skills/edgeone-makers-env-adaption/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "edgeone-makers-env-adaption", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
edgeone-makers-env-adaptionEnvironment-specific adaptation rules for EdgeOne Makers Skills running in sandboxed or restricted AI coding environments (e.g.
Edgeone Makers Env Adaption is an agent skill from TencentEdgeOne/edgeone-makers-tools. Environment-specific adaptation rules for EdgeOne Makers Skills running in sandboxed or restricted AI coding environments (e.g. WorkBuddy). Trigger when: the user is working in WorkBuddy or a sandboxed IDE where CLI prompts hang and network is proxy-isolated (but browser login still works). Covers: non-interactive CLI flags, network isolation workarounds, login in sandbox, proxy bypass, file preview constraints (MUST use http:// via dev server, NEVER file://, NEVER python -m http.server / npx serve), dev server…
Its SKILL.md is about 4.8k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It works with Python. The licence is MIT.
12 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit e6ead94. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Shell commands in SKILL.md call:
curlnpxpythonnpmFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md. Its commands use curl, npx and npm, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Names these keys or tokens, usually read from environment variables:
EDGEONE_PAGES_API_TOKENFrom names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Edgeone Makers Env Adaption loads about 4.8k tokens when it runs. Until then it costs about 140 tokens; SKILL.md has 2,254 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from TencentEdgeOne/edgeone-makers-tools at commit e6ead94, republished under its MIT licence (© TencentEdgeOne). 2,254 words, ~4,807 tokens.
.claude/skills/edgeone-makers-env-adaption/SKILL.md (or your agent's skills folder).This document describes the special constraints and adaptation rules for EdgeOne Makers Skills across different AI coding environments. Currently covered: WorkBuddy (Tencent sandboxed IDE)
When you reach the "display / preview" step, read this first before deciding how to call present_files:
┌─ Delivering finished work? ── Yes ──→ present_files(deployed EdgeOne URL) ✅
│
Enter ┤
preview │ ┌─ dev server running? ─ Yes ──→ present_files(http://127.0.0.1:8088/) ✅
└─ Still iterating? ───┤
└─ No ──→ start edgeone makers dev → present_files(...)| What you want to do | Correct approach | Wrong approach (breaks) |
|---|---|---|
| Preview local dev server | present_files("http://127.0.0.1:8088/") | ❌ Passing /path/to/index.html (IDE opens it via file://) |
| Preview a deployed project | present_files(deploy_url) with ?eo_token=... | ❌ Passing a local dist/index.html path |
| Start dev server | edgeone makers dev --name <p> --skip-env-sync | ❌ python -m http.server / npx serve |
| Verify dev server is up (agent-side API check) | curl --noproxy '*' http://127.0.0.1:8088/api/... ✅ works (same sandbox) | ❌ curl localhost:8088 / plain curl (proxy + IPv6 → 404/000) |
| Verify dev server is up (user-facing) | present_files(http://127.0.0.1:8088/) (platform tunnel) | ❌ Telling user to open 127.0.0.1:8088 — their browser can't reach the sandbox |
Core iron rule: inside a Makers project, any HTML / URL preview MUST go through the HTTP protocol. file:// looks convenient, but fetch / SSE / Blob / KV all break under it.
TypeError: Failed to fetch / CORS policy errorsWorkBuddy is a sandboxed remote IDE environment. When running AI coding tasks, it has the following constraints that differ from local development.
⚠️ Common misread: WorkBuddy is a desktop IDE sandbox, NOT a headless environment.
- CLI interactive prompts (site picker, confirmations) hang inside the sandbox → use the non-interactive flags in this section.
- BUT browser login (
edgeone login --site <x>) works fully — WorkBuddy launches the host OS browser AND routes the OAuth callback back into the sandbox.Deploy/link login should default to browser login and only fall back to
-t <token>after browser login is confirmed to fail. Do NOT jump to token login just because the caller is an "Agent".
Inside the WorkBuddy sandbox, CLI interactive prompts cause the process to hang forever. All edgeone CLI commands must carry non-interactive flags:
| Scenario | Required flag | Reason |
|---|---|---|
| Local development | --skip-env-sync | Skips the "sync environment variables?" confirmation |
| Linking a project | --name <project> | Skips the interactive project picker |
| Auth when not logged in | -t <token> | Passes the token directly, no login popup |
| Deploy output | --json | Machine-readable JSON, avoids ANSI parsing |
# Correct: local development
edgeone makers dev --name my-project --skip-env-sync
# Correct: deploy
edgeone makers deploy -n my-project --json
# Wrong: will hang
edgeone makers devToken resolution priority (the CLI checks in this order automatically):
-t <token> command-line argumentEDGEONE_PAGES_API_TOKEN environment variable<cwd>/.edgeone/auth.json (written by edgeone login --local)~/.edgeone/ global credentialsRecommended approach: browser login + the --local flag:
edgeone login --site china --local--local writes credentials to the project directory at <cwd>/.edgeone/auth.json, bypassing home-directory write restrictions.
Login status detection:
edgeone whoami # exit 0 = logged in, exit 1 = not logged in (does not hang)When is login actually required? Login is only needed when the project uses Blob or other credentialed backends — and strictly because of the dependency chain: Blob requires the project to be linked, and linking requires a logged-in account first. So edgeone makers dev for a pure-static site runs fine without login — do NOT force a login prompt for static-only previews. Login (or -t <token>) becomes mandatory the moment dev/deploy must touch Blob storage. (The trigger condition and the link chain live in makers-storage.)
CLI version requirement: >= 1.6.7 (older versions lack the non-interactive fixes; whoami will hang)
The Bash tool and edgeone makers dev run in the same sandbox (same machine), so the loopback dev server IS reachable from Bash. The earlier claim that "Bash curl is isolated and returns 404" is wrong — the failures are caused by the proxy (§5) and the IPv6 localhost (§4), not by network isolation.
| Verification method | Availability | Notes |
|---|---|---|
Built-in browser preview (present_files) | ✅ Available | Uses the platform tunnel; the ONLY way the user's browser can see the sandbox dev server |
| User's system terminal | ✅ Available | curl http://127.0.0.1:8088/ |
| Bash tool curl (agent-side API checks) | ✅ Available only with --noproxy '*' + 127.0.0.1 | Plain curl localhost:8088 fails: (a) localhost→::1 (§4), (b) proxy hijacks the request (§5) |
Practical rule:
curl --noproxy '*' http://127.0.0.1:8088/... from Bash to agent-side verify API endpoints during testing — this works (it was used to validate a full create→upload→like flow).127.0.0.1:8088 inside the sandbox; for a user-facing preview, pass the dev URL to present_files (platform tunnel) or deploy and share the live URL.The dev server listens on the IPv6 dual stack (::), but in the sandbox localhost resolves to ::1, causing false 404s.
# Correct
curl http://127.0.0.1:8088/
# Wrong (404 in the sandbox)
curl http://localhost:8088/The preview URL must also use 127.0.0.1:
present_files: http://127.0.0.1:8088/The sandbox injects an http_proxy environment variable; curl goes through the proxy by default, which swallows the SSE streaming response.
# Correct
curl --noproxy '*' http://127.0.0.1:8088/api/chat
# Wrong (returns "Empty reply" / status 000)
curl http://127.0.0.1:8088/api/chatThe built-in browser preview is not affected by the proxy.
The sandbox blocks writes to ~/.edgeone/, but allows writes to the project directory.
| Path | Writable | Notes |
|---|---|---|
<cwd>/.edgeone/ | ✅ | Where --local writes |
~/.edgeone/ | ❌ | EPERM error |
A setLocalData EPERM does not affect the running service; it only affects the CLI's local state persistence.
| Command | Execution mode | Reason |
|---|---|---|
npm install | Foreground sync | Usually 10-30s; running it in the background would leave later commands missing dependencies |
edgeone makers dev | Background async (run_in_background) | Long-running process, must not block the conversation |
edgeone makers deploy | Background async (run_in_background) | Cold deploys (build → upload → Process → live) routinely take 2–10+ minutes. The foreground wall-clock budget (~100s) SIGKILLs the CLI mid-deploy (exit 137) even while it keeps printing progress — you lose the final URL line. Run it in the background and wait for the completion notification. |
exit 137 = 128 + 9), regardless of whether it keeps printing. During a deploy the CLI prints Deployment in progress... elapsed: ~XXs every ~10s — those heartbeat lines do NOT reset or extend the budget. So "still printing → killed at ~100s" is expected, not a stall.Created deployment / Process has very likely continued on the server and gone live (confirmed in practice: a deploy killed at ~100s was later verified live, took ~674s end-to-end). Re-running deploy (same -n <project>) reuses the project and returns the URL.run_in_background: true for deploy so the CLI survives past the foreground budget and emits the final --json line with the live URL. Do not poll the task across turns — rely on the <task-notification> completion event. (Cross-turn TaskOutput may report the handle as "not found"; the process still finished.)--json still prints progress heartbeats to stdout/stderr, not a single clean JSON line. Parse the last line for the result object; treat the progress lines as noise. (This is a CLI cleanliness issue, unrelated to the kill behavior — do not assume "no output = process ended"; the sandbox judges liveness by OS process state, not by stdout bytes.)edgeone makers dev left running keeps holding port 8088 and may collide with the next dev/deploy. Stop it with TaskStop or kill it when the session moves on.After finishing development, start the dev server and preview directly — do not ask "do you want to preview?". Full flow:
⚠️ WorkBuddy default behavior: when you create an HTML file the platform may auto-open a preview via file:// — ignore it, that is not a valid preview. You must wait until
edgeone makers devis up, then re-open via the HTTP URL to override it.
edgeone makers dev --name <project> --skip-env-sync (background async, see §7)http://127.0.0.1:8088/ to present_files (note it is 127.0.0.1, not localhost — see §4)Only after the user confirms, run edgeone makers deploy -n <project> --json (background async, see §7 and §7.2 — it exceeds the foreground wall-clock budget).
Never pass a local HTML path to present_files — the IDE opens it via the file:// protocol, which breaks fetch / SSE / Blob / KV entirely. No exceptions, no "just a quick look" scenario.
# ✅ Correct
present_files("http://127.0.0.1:8088/") # dev server
present_files("https://my-app-w9t0lxe8.edgeone.cool?eo_token=...") # after deploy (full URL with query params)
# ❌ Wrong (IDE opens via file://, all APIs fail)
present_files("/Users/foo/dist/index.html")
present_files("./dist/index.html")
present_files("file:///Users/foo/dist/index.html")
# ❌ Wrong (truncated query params — the user gets a 401 when they open it)
present_files("https://my-app-w9t0lxe8.edgeone.cool") # missing ?eo_token=...⚠️ URL truncation = 401: the deploy URL's ?eo_token=...&eo_time=... are auth parameters; if truncated, the user gets a 401 UNAUTHORIZED when they open it. Every URL in your reply must include the full query string, including secondary references in tables, lists, and footnotes.
Violation symptoms (if you see these, immediately re-check § Quick Reference):
TypeError: Failed to fetch / Cross-Origin Request Blockedfetch / XMLHttpRequest calls fail@edgeone/pages-blob calls report Missing: deployCredential (even when the project is linked, because there is no HTTP context under file://)edgeone makers dev must NOT be replaced by any of the following:
python -m http.servernpx serve / npx http-serverhttp.createServer / expressReason: edgeone makers dev injects Blob credentials, emulates Cloud Functions routing, handles Edge Functions, and runs the middleware chain. A self-hosted server can only serve static files; its behavior diverges from production and produces mysterious "works locally, breaks on deploy" (or the reverse) issues.
The Next.js 15+ dev server trusts only localhost by default. Accessing it via 127.0.0.1 in the sandbox is treated as cross-origin, so the HMR WebSocket is blocked and the page becomes unresponsive.
You MUST add to next.config.js:
allowedDevOrigins: ["127.0.0.1"]Note: the value is a bare host, without an http:// prefix and without a port.
Projects that use Blob Storage or KV must ensure the project is linked (a .edgeone/project.json exists) before starting dev. When not linked, Blob/KV calls report Missing: deployCredential.
Precondition — login first: linking (and therefore Blob/KV) requires a logged-in account. The full chain is Blob → must be linked → linking requires login. If edgeone whoami returns exit 1, run edgeone login (browser) or use edgeone makers link -t <token> with a token before attempting to link. Do NOT try to link while unauthenticated.
Detect whether it is linked:
cat .edgeone/project.json 2>/dev/null && echo "LINKED" || echo "NOT LINKED"How to link:
# Explicit link (auto-creates the project if it does not exist)
edgeone makers link --name <project-name> -t <token>
# Or implicitly link via the dev command
edgeone makers dev --name <project-name> --skip-env-syncIf the project named by --name does not exist remotely, the link command creates it automatically.
Note: even for a pure static project, if the code imports @edgeone/pages-blob, not linking will always cause an error.
| Framework/package | Minimum version | Reason |
|---|---|---|
| EdgeOne CLI | >= 1.6.7 | Non-interactive fixes, whoami fail-fast, --json support |
EdgeOne CLI (anonymous deploy / claim only) | >= 1.6.29 | --anonymous and claim do not exist below this |
| Next.js | 16.x | The framework adapter tracks new versions |
| @edgeone/pages-blob | >= 0.0.14 | Older versions have known bugs |
Use create-next-app@latest rather than manually pinning an older version.
alert / confirm / prompt) don't work in WorkBuddy's built-in browserWorkBuddy's right-side preview panel does NOT render window.alert() / window.confirm() / window.prompt(). The call returns immediately without user interaction, so any handler gated on if (confirm("Delete?")) silently no-ops (a delete button appears to do nothing). The same page works fine in the user's real Chrome / Safari.
Rule: for any confirmation, prompt, or notification in the page, use an in-page custom modal (a <div> overlay with buttons wired via JS). Do NOT rely on the browser's built-in alert / confirm / prompt — the code looks correct in code review, works when the user opens the deployed URL in their own browser, and is silently broken in the WorkBuddy preview during dev/verification.
The user has no terminal in the WorkBuddy sandbox. Any command you print as an instruction is dead text — they cannot execute it, and a non-technical user cannot read it either.
| Situation | Do this | Not this |
|---|---|---|
| Anonymous project needs claiming | Run edgeone login --claim --local --json in the background and present its claimLoginUrl as a clickable link; the listener auto-links the project when the user finishes claiming | ❌ Printing edgeone makers claim --sid <token>, or presenting the bare claimUrl from the deploy JSON (the CLI never learns the login state) |
| The user needs to log in | Explain, then run edgeone login --site <china|global> --local yourself | ❌ "Run edgeone login in your terminal" |
| Something needs installing | Run it yourself in Bash | ❌ "Please run npm install -g edgeone" |
You are the one with shell access — use it. Only ask the user to type something when it genuinely cannot be automated (e.g. completing a login in the browser window you opened), and say plainly what you need from them.
WorkBuddy can collapse content emitted in the same message as a running or just-finished Bash command into a collapsible process entry. When the user asks to deploy, the deploy URL and claim link must never land there — if they do, the user sees nothing.
| Situation | Do this | Not this |
|---|---|---|
| A deploy just finished | Wait for the Bash call to fully return, then send the result (URL + claim link) as its own message | ❌ Emitting the result in the same message where the deploy ran, or attaching it to the command's stdout |
| Running a long command | Report progress only in later messages, after the tool call returns | ❌ Writing "here is your URL" inside the same turn as the command output |
© TencentEdgeOne, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Just SKILL.md in skills/edgeone-makers-tools/references/makers-env-adaption of TencentEdgeOne/edgeone-makers-tools.
Open the folder on GitHubat commit e6ead94
We found 2 copies of this SKILL.md (exact, near-identical or edited) in other folders, from 1 other GitHub owner. This page covers the copy in TencentEdgeOne/edgeone-makers-tools, which our catalogue first saw on October 7, 2026.
Edgeone Makers Env Adaption next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Edgeone Makers Env Adaption this skillTencentEdgeOne/edgeone-makers-tools | 1.9k | 1 repos | ~4.8k | Automated safety check: Pass | MIT | |
| MCP Server Builderanthropics/skills | 180k | 63 repos | ~2.3k | Automated safety check: Pass | Apache-2.0 | |
| PDF Processinganthropics/skills | 180k | 48 repos | ~2k | Automated safety check: Pass | Proprietary | |
| NotebookLM Research AssistantPleasePrompto/notebooklm-skill | 7.8k | 14 repos | ~2.4k | Automated safety check: Notes | MIT | |
| Manim Video Productionbrowser-use/video-use | 28k | 6 repos | ~3k | Automated safety check: Pass | MIT | |
| Code Review ChecklistshareAI-lab/learn-claude-code | 78k | 5 repos | ~1.1k | Automated safety check: Pass | MIT |
anthropics/skills
Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.
anthropics/skills
Handles everyday PDF jobs in Python and on the command line: extract text and tables, merge, split, rotate, watermark, fill forms, encrypt and OCR.
PleasePrompto/notebooklm-skill
Lets Claude Code ask questions of your Google NotebookLM notebooks through browser automation and return answers grounded in your uploaded sources.
browser-use/video-use
Produces math and technical explainer videos with Manim Community Edition: concept animations, equation derivations, algorithm walkthroughs and data stories.
shareAI-lab/learn-claude-code
Reviews code against a five-part checklist covering security, correctness, performance, maintainability and testing, and reports findings in a fixed format.
hugohe3/ppt-master
Generates editable PowerPoint decks, rebuilds slides from images, fills .pptx templates and polishes existing presentations through routed workflows.
TencentEdgeOne/edgeone-makers-tools
Migrate existing AI agent projects (LangChain, LangGraph, OpenAI Agents SDK, Claude Agent SDK, CrewAI) to EdgeOne Makers platform conventions.
TencentEdgeOne/edgeone-makers-tools
This skill guides building AI agent endpoints on EdgeOne Makers — five framework routes (DeepAgents, LangGraph, CrewAI, OpenAI Agents SDK, Claude Agent SDK), platform-injected context.store /…
TencentEdgeOne/edgeone-makers-tools
EdgeOne Makers CLI command reference. An agent skill from TencentEdgeOne/edgeone-makers-tools.
TencentEdgeOne/edgeone-makers-tools
Project structure templates and scaffolding recipes for typical EdgeOne Makers applications — full-stack apps, static sites, API services, and AI agent projects.
TencentEdgeOne/edgeone-makers-tools
KV and Blob storage services on EdgeOne Makers. An agent skill from TencentEdgeOne/edgeone-makers-tools.
TencentEdgeOne/edgeone-makers-tools
V8-based lightweight edge functions on EdgeOne Makers. An agent skill from TencentEdgeOne/edgeone-makers-tools.
Works with
Environment-specific adaptation rules for EdgeOne Makers Skills running in sandboxed or restricted AI coding environments (e.g. Edgeone Makers Env Adaption is an agent skill from TencentEdgeOne/edgeone-makers-tools.g.
Edgeone Makers Env Adaption fits situations like: : the user is working in WorkBuddy; A sandboxed IDE where CLI prompts hang and network is proxy-isolated (but browser login still works).
Run `npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a claude-code`. Or copy the skill folder (skills/edgeone-makers-tools/references/makers-env-adaption in TencentEdgeOne/edgeone-makers-tools) into .claude/skills/edgeone-makers-env-adaption in your project. Claude Code loads it when a task matches its description.
Run `npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a codex`. Or copy the skill folder (skills/edgeone-makers-tools/references/makers-env-adaption in TencentEdgeOne/edgeone-makers-tools) into .agents/skills/edgeone-makers-env-adaption in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add TencentEdgeOne/edgeone-makers-tools --skill edgeone-makers-env-adaption -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/edgeone-makers-env-adaption, .gemini/skills/edgeone-makers-env-adaption, .github/skills/edgeone-makers-env-adaption and .opencode/skills/edgeone-makers-env-adaption in your project.
Going by SKILL.md and its folder, Edgeone Makers Env Adaption needs the command-line tools its instructions call (curl, npx, python and npm) and credentials named EDGEONE_PAGES_API_TOKEN. Our summary lists: Python 3; Node.js; A credential in EDGEONE_PAGES_API_TOKEN.
SKILL.md contains no URLs. Its commands use curl, npx and npm, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Edgeone Makers Env Adaption is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 4.8k tokens (SKILL.md is roughly 19k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Edgeone Makers Env Adaption: MCP Server Builder (anthropics/skills, 180k stars), PDF Processing (anthropics/skills, 180k stars), NotebookLM Research Assistant (PleasePrompto/notebooklm-skill, 7.8k stars) and Manim Video Production (browser-use/video-use, 28k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
TencentEdgeOne (a GitHub organization) maintains it in TencentEdgeOne/edgeone-makers-tools, which has 1,861 GitHub stars. The repository holds 12 skills in this directory. The repository was last updated on October 8, 2026.
Source: TencentEdgeOne/edgeone-makers-tools on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.