What does this skill do, and when should you use it?
This skill is a smart search router from the OpenCLI collection. It does not memorize commands; instead, it dynamically discovers available sites and subcommands via `opencli list -f yaml` and `opencli <site> -h`, then routes user queries to the most appropriate sources (AI or vertical). The skill enforces per-question budgets and rate limits to avoid over-calling, and requires a search summary after each query. It targets AI-agent users needing efficient, reliable search, especially for Chinese content.
The skill performs these operations: first runs opencli list -f yaml to get the live registry and confirm candidate sites; then runs opencli <site> -h to see subcommands, and opencli <site> <command> -h for parameters if needed; selects one AI source (grok, doubao, gemini) based on user specification or language context, or chooses specialized sources; updates a call ledger after each search recording site, query, count, and status; at query end, appends a fixed-format 'search summary' section listing sites used, query terms, and call counts, and notes any rate-limited skips.
- A developer searching for Chinese technical documentation uses doubao for Chinese-context answers and supplements with bilibili or zhihu for original content.
- A product manager comparing Twitter trends uses grok for English sentiment and calls the twitter command to fetch raw tweets.
- A job seeker looking for specific roles uses gemini for global job search and supplements with linkedin's job-detail command for detailed postings.
- A user researching shopping options uses doubao for Chinese recommendations and calls amazon or 1688 commands for raw product results.
- A researcher needing authoritative academic sources uses gemini for background synthesis and supplements with google-scholar for original papers.
How do you install this skill?
- This skill depends on external opencli tool; its security is out of scope here. Ensure opencli is trusted and audited.
- The skill lacks explicit user confirmation for sensitive operations; use with caution.
- Some recommended sites (e.g., Google, Twitter/X) may be unreachable from mainland China; verify network access.
- Shell / CLI
- Local filesystem
Node.jsOpenCLI
This skill is part of the OpenCLI collection. To install: first install OpenCLI (recommended via the OpenCLIApp desktop app or npm global install npm install -g @jackwener/opencli), then install the browser bridge extension. The skill itself is installed individually with npx skills add jackwener/opencli --skill smart-search.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/jackwener/OpenCLI.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/smart-search" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Ensure OpenCLI is installed and run opencli doctor to verify connectivity. In your AI agent, the skill triggers automatically when the user asks to search, query, or research. Before every use, perform the mandatory preflight: opencli list -f yaml and opencli <site> -h. Then select the data source based on routing rules, execute search commands, and update the call ledger. After the query, append a fixed-format 'search summary' to your response.
What are this skill's strengths and limitations?
- Avoids documentation drift by relying on live command help.
- Strict rate limits and budget control prevent overuse and save resources.
- Mandatory search summary increases transparency and auditability.
- Supports Chinese context with built-in Chinese AI sources.
- Requires full OpenCLI installation and browser connection, making initial setup complex.
- Rate limits may lead to insufficient information; gaps must be accepted.
- No test suite or examples provided; actual effectiveness requires self-validation.
- Only covers sites supported by OpenCLI, limiting vertical reach.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| OpenCLI Smart Search Router this page | 41 · Not recommended | ★ 30k | 17d ago | Apache-2.0 |
| OpenCLI AutoFix — Automatic Adapter Repair | 58 · Recommended | ★ 30k | 17d ago | Apache-2.0 |
| playwright-cli Browser Automation Skill | 56 · Use with care | ★ 11k | 13d ago | MIT |
| Playwright Browser Automation ✓ Microsoft · Official | 46 · Use with care | ★ 97k | 3d ago | Apache-2.0 |
| OpenCLI Adapter Authoring Guide | 56 · Use with care | ★ 30k | 17d ago | Apache-2.0 |
How did FollowSkills review this skill?
Skill is a router and does not directly execute high-risk operations, but relies on external opencli tool which involves browser login state and downloads. No explicit data flow, user confirmation or sensitive info handling; no rollback. No malicious behavior found, but permission boundaries and user confirmation missing, so deducted.
Skill doc is self-consistent with mandatory pre-checks and live help, but lacks examples of error handling or how to feedback on abnormal input. Relies on external opencli commands which may fail due to network/unavailability, but fallback strategy provided. Static assessment cannot verify execution, so <=10.
Clear scenario for Chinese users with Chinese sources and query advice, but no explicit non-fit scope. Some overseas sites may be unreachable in mainland, but fallback provided. Trigger conditions clear but boundaries and precision could be improved.
Well-structured with layered reference files, but no version history or clear maintenance responsibility. License/install visible at repo level, but skill lacks troubleshooting or update path. Few examples, no known limitations.
Design rationale sound with live-help routing to avoid drift, but no sample outputs or verification results. Static cannot confirm output quality and depends on external commands, so <=7.
Repo has CI and tests, but not covering this skill's routing logic. No executable test cases in the skill. Static cannot reproduce, so <=5.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →