$npx -y skills add parallel-web/parallel-agent-skills --skill parallel-web-searchDEFAULT for all research and web queries. Use for any lookup, research, investigation, or question needing current info. Fast and cost-effective. Only use parallel-deep-research if user explicitly requests 'deep' or 'exhaustive' research.
| 1 | # Web Search |
| 2 | |
| 3 | Search the web for: $ARGUMENTS |
| 4 | |
| 5 | ## Command |
| 6 | |
| 7 | Choose a short, descriptive filename based on the query (e.g., `ai-chip-news`, `react-vs-vue`). Use lowercase with hyphens, no spaces. Substitute it into the command **inline** — `$FILENAME` and `<keyword>` below are placeholders, not shell variables; do not copy them verbatim. |
| 8 | |
| 9 | ```bash |
| 10 | parallel-cli search "$ARGUMENTS" -q "<keyword1>" -q "<keyword2>" --json --max-results 10 --excerpt-max-chars-total 27000 -o "/tmp/$FILENAME.json" |
| 11 | ``` |
| 12 | |
| 13 | Concrete example for a query about React 19: |
| 14 | |
| 15 | ```bash |
| 16 | parallel-cli search "latest React 19 features and adoption" -q "React 19" -q "concurrent rendering" --json --max-results 10 --excerpt-max-chars-total 27000 -o "/tmp/react-19-features.json" |
| 17 | ``` |
| 18 | |
| 19 | The first argument is the **objective** — a natural language description of what you're looking for. It replaces multiple keyword searches with a single call for broad or complex queries. Add `-q` flags for specific keyword queries to supplement the objective. The `-o` flag saves the full results to a JSON file for follow-up questions. |
| 20 | |
| 21 | Options if needed: |
| 22 | |
| 23 | - `--after-date YYYY-MM-DD` for time-sensitive queries |
| 24 | - `--include-domains domain1.com,domain2.com` to limit to specific sources |
| 25 | - `--exclude-domains domain.com` to filter out noisy sources |
| 26 | - `--mode turbo` for simple fact lookups where speed and cost matter most (p50 ~200ms, lowest cost). English and Japanese queries only |
| 27 | - `--mode advanced` for harder questions (multi-step, agentic search). Default `basic` is right for almost everything; escalate to `advanced` only when basic results are insufficient, and drop to `turbo` for high-volume simple lookups |
| 28 | - `--location us` (ISO 3166-1 alpha-2) for geo-targeted results |
| 29 | |
| 30 | ## Parsing results |
| 31 | |
| 32 | Do not set `max_output_tokens` on the command execution — the output is already bounded by `--max-results` and `--excerpt-max-chars-total`. Capping output tokens will truncate the JSON and break parsing. |
| 33 | |
| 34 | **Prefer reading from the saved `-o` file**, not stdout. Even bounded output regularly exceeds harness stdout limits and gets truncated. Read `/tmp/$FILENAME.json` for the authoritative payload. For each result, extract: |
| 35 | |
| 36 | - title, url, publish_date |
| 37 | - Useful content from excerpts (skip navigation noise like menus, footers, "Skip to content") |
| 38 | |
| 39 | ## Response format |
| 40 | |
| 41 | **CRITICAL: Every claim must have an inline citation.** Use markdown links like [Title](URL) pulling only from the JSON output. Never invent or guess URLs. |
| 42 | |
| 43 | Synthesize a response that: |
| 44 | |
| 45 | - Leads with the key answer/finding |
| 46 | - Includes specific facts, names, numbers, dates |
| 47 | - Cites every fact inline as [Source Title](url) — do not leave any claim uncited |
| 48 | - Organizes by theme if multiple topics |
| 49 | |
| 50 | **End with a Sources section** listing every URL referenced: |
| 51 | |
| 52 | ```text |
| 53 | Sources: |
| 54 | - [Source Title](https://example.com/article) (Feb 2026) |
| 55 | - [Another Source](https://example.com/other) (Jan 2026) |
| 56 | ``` |
| 57 | |
| 58 | This Sources section is mandatory. Do not omit it. |
| 59 | |
| 60 | After the Sources section, mention the output file path (`/tmp/$FILENAME.json`) so the user knows it's available for follow-up questions. |
| 61 | |
| 62 | ## Setup |
| 63 | |
| 64 | If `parallel-cli` is not found, install and authenticate: |
| 65 | |
| 66 | ```bash |
| 67 | /parallel:parallel-cli-setup |
| 68 | ``` |
| 69 | |
| 70 | If `parallel-cli search` returns `403`, tell the user balance is likely required. Offer to run `parallel-cli balance get`, and if needed ask for explicit confirmation before running `parallel-cli balance add <amount_cents>`. Then retry the original search command. |