Metaphor

@Metaphor fetch the full text and metadata for documents ID-7829, ID-7830, and ID-7831 so we can compare their methodology sections.

Retrieved 3 documents. Document ID-7829 (published March 2024) describes a longitudinal cohort study with 1,200 participants tracked over 18 months. Document ID-7830 (June 2023) uses a cross-sectional survey design with stratified sampling across 8 regions. Document ID-7831 (November 2023) employs a mixed-methods approach combining qualitative interviews with quantitative analysis of administrative records. All three include detailed methodology sections with sample size justifications and statistical power calculations.

This example shows Metaphor's document retrieval in action. The MCP returns parsed HTML or plain text plus metadata, which the AI can then analyze or summarize. Requires valid document IDs from a prior search or known corpus. Rate limits depend on your API tier—batch requests like this count as multiple calls.

@Metaphor pull the content for document ID-4512 and extract all the key findings into a bulleted list for our research brief.

Document ID-4512 retrieved successfully. Key findings: • Remote work adoption increased productivity by 13% in knowledge-worker roles but decreased collaboration frequency by 22%. • Hybrid schedules (3 days in-office) showed the highest employee satisfaction scores. • Onboarding effectiveness dropped 31% for fully remote new hires without structured mentorship. • Tool sprawl (average 11 SaaS apps per employee) correlated with reported communication friction. • Teams with async-first norms reported 18% fewer meeting hours per week.

Demonstrates pairing Metaphor's retrieval with the AI's synthesis capability. The MCP fetches raw document text; the AI extracts and formats insights. This workflow is ideal for turning long-form content into actionable summaries. Ensure your API key has access to the document corpus you're querying.

@Metaphor get the full text for documents ID-9001 through ID-9005 and identify which one discusses supply chain risk mitigation strategies in the most detail.

Fetched 5 documents. Document ID-9003 provides the most comprehensive coverage of supply chain risk mitigation, dedicating 4 sections (approximately 3,200 words) to strategies including supplier diversification, inventory buffering, demand forecasting models, and geopolitical risk assessment frameworks. It includes 7 case studies from automotive and electronics industries. Documents ID-9001 and ID-9004 mention supply chain risk briefly in passing. ID-9002 focuses on financial hedging rather than operational strategies. ID-9005 covers logistics optimization but not risk mitigation specifically.

Shows Metaphor handling batch retrieval for comparative analysis. The AI reads across multiple documents to answer a research question. Useful when you have a set of candidate documents and need to identify the most relevant one. Be mindful of token costs—5 full documents can consume significant context window depending on length.

When Metaphor beats manual research for feature validation

A 6-person product team at a B2B SaaS company needs to validate whether competitors have shipped a feature they're considering. The PM pastes 3-5 competitor URLs into Switchy, and the Metaphor MCP pulls full text and metadata in one call. This works when you already know the URLs—Metaphor's strength is bulk retrieval, not discovery. If your workflow is 'find articles about X topic,' you need a search MCP instead. But if you're tracking a known set of competitor blogs, changelog pages, or landing pages, Metaphor turns a 20-minute tab-switching session into a 90-second batch pull. The API key setup takes under 5 minutes. Worth it if you run this drill weekly or more.

Metaphor for pulling external docs into support threads

A 3-person support team at a dev tools startup fields questions that reference third-party API docs. They maintain a list of 15-20 external documentation URLs (Stripe, Twilio, AWS) that customers ask about. When a ticket comes in, the agent drops the relevant doc URL into Switchy, and Metaphor fetches the current version of that page. This beats copy-pasting because the MCP grabs structured metadata (publish date, title) alongside the text, so the agent can confirm they're citing the latest version. The trade-off: Metaphor only retrieves by ID, so you need the URL upfront. If you need to search across those docs, you're back to manual. Best for teams with a stable, curated list of external references.

When Metaphor speeds up content curation by 40%

A 2-person marketing team publishes a weekly industry newsletter. They maintain a spreadsheet of 30 blog URLs (analyst firms, trade publications, niche Substacks) and want to pull the latest post from each every Monday. They paste the URL list into Switchy, and Metaphor returns full text and metadata in one batch. The team scans the output, picks 5-7 pieces to summarize, and ships the newsletter by noon. This works because the source list is stable and the team wants full text, not search results. If your sources change weekly or you need to filter by topic before retrieval, Metaphor adds friction. But for a fixed rotation of trusted sources, it cuts research time from 90 minutes to 50.