agentleFS
PeiiiiPeiiii/nextclaw/skills/opencli/SKILL.md

opencli

Skill263 starsChanged 35 days agoMITView on GitHub

Use when the user wants to use websites, browser login sessions, Electron apps, or external CLIs through a local OpenCLI setup, especially when setup guidance, readiness checks, and safe task execution are needed.

  • Installs packages

10 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

29 rules. 0 are common, 21 are distinctive, and 8 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 21

    Distinctive

    This author's own choices. Worth a look.

  • 8

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use this skill to help the user work through a local opencli installation from inside NextClaw.

    line 10
  2. Distinctive:

    Do not pretend the environment is ready when it is not.

    line 24
  3. Very specific:

    When the user asks for an opencli-powered task, follow this order.

    line 44
  4. Very specific:

    If the task does not fit any supported opencli shape, say so clearly.

    line 55
  5. Very specific:

    If missing, explain that opencli must be installed locally first.

    line 65
  6. Distinctive:

    After installation, continue with readiness checks instead of jumping straight into the user task.

    line 73
  7. Distinctive:

    Use this as the main readiness gate.

    line 83
  8. Very specific:

    If doctor fails, do not proceed to the real task yet.

    line 85
  9. Distinctive:

    Do not assume desktop adapters are universally cross-platform.

    line 99
  10. Distinctive:

    Verify the target command exists before assuming passthrough is ready.

    line 103
  11. Distinctive:

    Do not rely on that behavior silently.

    line 111
  12. Distinctive:

    If the underlying CLI is missing, explain it and ask for explicit user confirmation before any install path.

    line 112
  13. Distinctive:

    When support is uncertain, run:

    line 116
  14. Very specific:

    Only use commands that can be confirmed from the installed opencli command list.

    line 130
  15. Distinctive:

    Prefer read-only commands before write commands.

    line 134
  16. Distinctive:

    If the task is exploratory, start with a lightweight read command or status command before a more invasive action.

    line 136
  17. Distinctive:

    If the user asks for an unsupported site or command, say it is unsupported instead of improvising.

    line 137
  18. Very specific:

    If opencli doctor or command output indicates missing login, missing extension, or missing app state, stop and guide the user.

    line 138
  19. Distinctive:

    Do not bury setup details.

    line 139
  20. Distinctive:

    When browser setup is missing, guide the user through this exact idea:

    line 143
  21. Distinctive:

    Make sure Chrome stays running.

    line 149
  22. Distinctive:

    Do not claim browser commands are safe to run before these checks pass.

    line 157
  23. Distinctive:

    When a user is sensitive to privacy or account risk, explain these points clearly:

    line 161
  24. Distinctive:

    Do not frame this as "no-risk magic".

    line 168
  25. Distinctive:

    Use patterns like these:

    line 172
  26. Very specific:

    If the user wants a specific command and you are unsure whether it exists, check opencli list first.

    line 184
  27. Distinctive:

    Do not continue to the real task.

    line 196
  28. Distinctive:

    Ask the user to log into the target site in Chrome and try again.

    line 202
  29. Distinctive:

    Ask for explicit confirmation before proceeding.

    line 213

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

OpenCLI

Overview

Use this skill to help the user work through a local opencli installation from inside NextClaw.

This skill is intentionally decoupled:

  • The skill owns explanation, onboarding, readiness checks, workflow choice, and risk handling.
  • opencli owns actual execution.

From the user's point of view, the experience should feel complete:

  • install or update opencli if needed,
  • guide the user through browser or app setup,
  • verify readiness,
  • then run the real task.

Do not pretend the environment is ready when it is not.

What This Skill Covers

  • public read-only commands,
  • browser tasks that reuse the user's Chrome login session,
  • Electron or desktop app adapters that opencli already supports,
  • passthrough to external CLIs through opencli,
  • first-use setup and troubleshooting.

What This Skill Does Not Cover

  • inventing commands that do not appear in opencli list,
  • claiming support for a site or app that opencli does not actually expose,
  • silently bypassing missing setup,
  • silently triggering destructive or write actions,
  • treating third-party runtime behavior as native NextClaw behavior.

First-Use Workflow

When the user asks for an opencli-powered task, follow this order.

1. Classify the task

Classify the task into one of these:

  • public read task,
  • browser session task,
  • desktop app task,
  • external CLI passthrough task.

If the task does not fit any supported opencli shape, say so clearly.

2. Check whether opencli exists

Run:

command -v opencli

If missing, explain that opencli must be installed locally first.

Recommended install:

npm install -g @jackwener/opencli

After installation, continue with readiness checks instead of jumping straight into the user task.

3. Run readiness check

Run:

opencli doctor

Use this as the main readiness gate.

If doctor fails, do not proceed to the real task yet. Diagnose the failure first.

4. Apply task-specific prerequisites

For browser session tasks:

  • Chrome must be running.
  • The user must already be logged into the target website in Chrome.
  • The OpenCLI Browser Bridge extension must be installed and enabled.

For desktop app tasks:

  • The target app must already be installed.
  • The target app should be open before execution when appropriate.
  • Do not assume desktop adapters are universally cross-platform; verify command availability first.

For external CLI passthrough tasks:

  • Verify the target command exists before assuming passthrough is ready.
  • Example:
command -v gh
  • opencli documentation says some passthrough commands may attempt auto-install of missing CLIs.
  • Do not rely on that behavior silently.
  • If the underlying CLI is missing, explain it and ask for explicit user confirmation before any install path.

5. Discover supported commands

When support is uncertain, run:

opencli list

Prefer:

opencli list -f yaml

when structured inspection is easier.

Only use commands that can be confirmed from the installed opencli command list.

Safe Execution Rules

  • Prefer read-only commands before write commands.
  • For writes, posting, deleting, following, publishing, or account-affecting actions, ask for explicit confirmation first.
  • If the task is exploratory, start with a lightweight read command or status command before a more invasive action.
  • If the user asks for an unsupported site or command, say it is unsupported instead of improvising.
  • If opencli doctor or command output indicates missing login, missing extension, or missing app state, stop and guide the user.
  • Do not bury setup details. Explain the missing prerequisite directly.

Browser Setup Guidance

When browser setup is missing, guide the user through this exact idea:

  1. Install the latest OpenCLI browser extension from the upstream releases page.
  2. Open chrome://extensions.
  3. Turn on Developer mode.
  4. Load the unpacked extension or the unzipped release bundle.
  5. Make sure Chrome stays running.
  6. Log into the target site in Chrome.
  7. Re-run:
opencli doctor

Do not claim browser commands are safe to run before these checks pass.

Privacy And Trust Guidance

When a user is sensitive to privacy or account risk, explain these points clearly:

  • opencli reuses the user's existing Chrome login state.
  • Its browser bridge communicates through a local daemon on localhost.
  • The extension requests browser permissions such as debugger, tabs, cookies, activeTab, and alarms.
  • The skill should surface this clearly before high-trust tasks.

Do not frame this as "no-risk magic". Frame it as a local tool with explicit browser permissions and explicit account reuse.

Recommended Command Patterns

Use patterns like these:

opencli list
opencli doctor
opencli hackernews top --limit 5
opencli reddit search "AI"
opencli youtube transcript "<url>"
opencli bilibili hot --limit 5
opencli gh pr list --limit 5

If the user wants a specific command and you are unsure whether it exists, check opencli list first.

Troubleshooting

opencli not found

  • Explain that the local CLI is not installed.
  • Guide the user through installation.
  • Re-check with command -v opencli.

doctor reports extension or daemon problems

  • Do not continue to the real task.
  • Guide the user through extension setup and rerun opencli doctor.

Browser command returns empty data or unauthorized behavior

  • Explain that the most likely cause is missing or expired Chrome login state.
  • Ask the user to log into the target site in Chrome and try again.

Desktop command fails

  • Confirm the target app is actually supported.
  • Confirm the app is open.
  • Confirm the command exists in opencli list.

External CLI passthrough may install another CLI

  • Explain that this may modify the local machine.
  • Ask for explicit confirmation before proceeding.

Success Criteria

The skill is working correctly when:

  • the user understands what opencli is being used for,
  • missing setup is identified before task execution,
  • readiness is checked with opencli doctor,
  • the user is guided through any required setup,
  • supported commands are confirmed via opencli list when needed,
  • and the final task runs only after the environment is truly ready.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.