Skip to content
cccskills
FreePublished on GitHub

web-search

Searches the open web for docs, comparisons, and news via You.com MCP or WebSearch. Use when a topic needs knowledge beyond GitHub, forums, or academia.

The English summary is being prepared. Showing the original description.

Works withGitHub

Run score Not tested

See how to install

Files included (1)

  • SKILL.md3.2 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Web Search

When To Use

  • Finding vendor documentation, standards, or official guides
  • Comparisons and benchmarks across products or approaches
  • News, release notes, and recent status for a technology
  • Standalone searches, or as one channel of a /tome:research session: the planner dispatches tome:web-searcher from medium depth, and the channel runs a positive control (RFC 2119) before its topic queries so its silence can count in the coverage verdict

When NOT To Use

  • Code implementations (use Skill(tome:code-search))
  • Community opinions (use Skill(tome:discourse))
  • Academic literature (use Skill(tome:papers))

Search the open web for pages about a topic. This is the general-knowledge channel: the pages no domain-specific channel targets.

Sources (Priority Order)

  1. You.com MCP server (you-search tool) when configured
  2. WebSearch tool as fallback (no configuration needed)

The You.com MCP server is optional. Without it the channel runs entirely on the built-in WebSearch tool, so a session never fails because the server is absent.

Setup (Optional)

To prefer You.com results, add the MCP server to your Claude Code config once:

# Keyless free profile
claude mcp add you --transport http https://api.you.com/mcp?profile=free

# Authenticated (higher rate limits; YDC_API_KEY from you.com)
claude mcp add you --transport http https://api.you.com/mcp \
  --header "Authorization: Bearer $YDC_API_KEY"

A key is only needed for the authenticated endpoint. The free profile needs no key and no account.

Workflow

  1. Build search queries using tome.channels.web.expand_web_queries()
  2. Run each query through the You.com MCP you-search tool if the server is configured, otherwise through WebSearch
  3. Parse results via parse_you_mcp_result() (raises ValueError on a response shape that is not a you-search envelope) or parse_websearch_result() (raises ValueError on a result with no URL)
  4. Rank via rank_web_findings()
  5. Return Finding objects

Failure Handling

  • If the You.com MCP tool errors or is unreachable, retry the query through WebSearch and record the fallback in the query log so the report can show the channel ran degraded rather than ok
  • If both paths fail for every query, report the failure explicitly rather than returning fabricated or empty findings

Exit Criteria

  • At least one web search query built and executed for the requested topic, through You.com MCP or WebSearch
  • Results parsed into Finding objects with title, URL, and a relevance score, marked channel="web"
  • Findings returned from rank_web_findings() sorted by relevance score descending
  • If the You.com path failed and WebSearch was used, the fallback is reported rather than passed off as the primary source
  • If no results are found for any query, this is reported explicitly rather than returning an empty list silently

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Shapes turns action-first: action leads, steps numbered, state restated. Use for ADHD-friendly output. Do not use to trim tokens; use response-compression.

The English summary is being prepared. Showing the original description.

athola/claude-night-market343Updated Oct 24, 2026

Inverts burden of proof for code additions. Use when reviewing PRs, planning refactors, or running unbloat to challenge every addition's necessity.

The English summary is being prepared. Showing the original description.

athola/claude-night-market343Updated Oct 24, 2026

Tracks per-agent token usage and flags waste in parallel dispatch. Use when evaluating parallel agent efficiency or after a multi-agent run.

The English summary is being prepared. Showing the original description.

athola/claude-night-market343Updated Oct 24, 2026

Coordinates Claude agent teams via filesystem protocol. Use when orchestrating parallel agents with task dependencies. Do not use for single-agent tasks.

The English summary is being prepared. Showing the original description.

athola/claude-night-market343Updated Oct 24, 2026

Evaluates API surface design, consistency, and exemplar alignment. Use when reviewing public API changes or before releasing a new API surface.

The English summary is being prepared. Showing the original description.

athola/claude-night-market343Updated Oct 24, 2026

Selects architecture paradigm via research before scaffolding. Use when architecture is undecided and the choice needs justification and documentation.

The English summary is being prepared. Showing the original description.

athola/claude-night-market343Updated Oct 24, 2026

See all skills by athola

Report a problem with this skill