upstash/context7 releases

Patch Changes

• d0e4a48: Create a fresh McpServer per HTTP request. Sharing one across requests let any concurrent transport.close clear the shared Protocol._transport, which broke sendNotification for in-flight long-running tool calls.
• 1aa3430: Remove research mode entirely from the MCP server and CLI. The query-docs MCP tool no longer accepts or forwards a researchMode parameter, and the CLI no longer exposes a --research flag on ctx7 docs.

Patch Changes

• 1aa3430: Remove research mode entirely from the MCP server and CLI. The query-docs MCP tool no longer accepts or forwards a researchMode parameter, and the CLI no longer exposes a --research flag on ctx7 docs.

Patch Changes

• 772da3a: Stream MCP tool responses over SSE so HTTP headers flush before client fetch timeouts. Switching enableJsonResponse to false makes the SDK return the HTTP response synchronously after request validation, so headers are sent in milliseconds instead of being buffered until the tool completes. This fixes clients that cap the underlying fetch waiting for headers (e.g., Claude Code's 60s wrapFetchWithTimeout).

Patch Changes

• 8274bd0: Add missing tool annotations
• ff6c1be: Remove the researchMode parameter from the query-docs tool's input schema. The underlying API still supports research mode, but several MCP clients hit per-request timeouts (60s defaults) on long-running research calls in ways that can't always be solved server-side. Hiding the parameter prevents agents from invoking it through MCP until the timeout story is reliable across clients.

Patch Changes

• 1b0c211: Add endpoint for OpenAI Apps SDK domain verification.

Minor Changes

• 17b864f: Expose research mode through the MCP researchMode tool and the CLI docs --research flag for deep, agent-driven documentation answers.

Minor Changes

• 17b864f: Expose research mode through the MCP researchMode tool and the CLI docs --research flag for deep, agent-driven documentation answers.

Patch Changes

• 4feee15: Add CLI update notifications and a new ctx7 upgrade command. The CLI now checks for newer versions with cached state, shows a non-blocking notice before interactive commands, and provides safer upgrade guidance across npm, pnpm, bun, and ephemeral runner setups.
• f056b14: Add ctx7 remove as the cleanup counterpart to ctx7 setup, with safer detection and removal behavior. The command now prompts only for agents with actual Context7 artifacts, preserves non-Context7 MCP configuration when removing entries, and includes stronger test coverage for JSON and TOML cleanup.

Patch Changes

• 3f6e310: Fix skill installation path validation on Windows so valid files inside the target directory are not rejected due to backslash-separated resolved paths.

Patch Changes

• 00833f9: Preserve Node's default trusted CAs when NODE_EXTRA_CA_CERTS is configured, and add a regression test for custom CA loading.

Patch Changes

• 33f2338: Add Codex-specific CLI setup guidance so generated rules and the installed find-docs skill tell Codex to rerun Context7 CLI requests outside the default sandbox after DNS or network failures.

Patch Changes

• bc8eaf1: Add --all-agents and --yes support to ctx7 skills install for non-interactive multi-agent installs.

Patch Changes

• 8322879: Improve resolve libryar id tool prompt to provide the libraryName query with proper format

Patch Changes

• fb29170: Add Gemini CLI support to setup command
• 89d4862: Use GITHUB_TOKEN/GH_TOKEN or gh CLI auth for skill downloads to avoid GitHub API rate limits and support private repos
• 8322879: Improve resolve libryar id tool prompt to provide the libraryName query with proper format

Patch Changes

• 658ec67: Add --version/-v flag to MCP CLI
• 8322879: Improve resolve libryar id tool prompt to provide the libraryName query with proper format

Patch Changes

• 6961bdd: Allow re-selecting already configured agents in ctx7 setup and overwrite existing MCP config entries instead of skipping them. Fix TOML replacement to correctly handle sub-sections and prevent whitespace drift on repeated runs.

Patch Changes

• a667712: Update search filter warning
• be1a39a: Update server metadata and instructions.

Patch Changes

• a667712: Update search filter warning

• d739f9b: Fix OpenCode MCP setup to resolve all config file variants (opencode.json, opencode.jsonc, .opencode.json, .opencode.jsonc)

• 4f13168: Install rules alongside skills in ctx7 setup for better trigger rates

  • CLI setup now installs a rule file for each agent (previously only installed the skill)
  • Rule content fetched from GitHub, with agent-specific formatting (alwaysApply for Cursor)
  • Updated find-docs skill description for higher invocation rates (66% -> 98%)
  • Added Codex agent support with AGENTS.md append
  • OpenCode now writes to AGENTS.md instead of .opencode/rules/
  • Selective rule content with explicit when-to-use/when-not-to-use guidance

• c3c2647: Use ~/.agents/skills instead of ~/.config/agents/skills for global universal skill installs

Patch Changes

• fae6127: Add active teamspace name to whoami command output
• 4b63117: Reorder setup mode choices to show MCP server first
• 18b3292: Add token refresh support, centralize auth constants, switch whoami to internal API endpoint with teamspace display, and add unit tests for CLI auth utilities and commands

Patch Changes

• 7e60d05: - feat(cli): track install count events when skills are installed via ctx7 setup

Patch Changes

• 62dc278: - feat(cli): enumerate popularity with a 4-star scale in skill search, install, and suggest results
  • feat(cli): show install count range and trust score in skill hover details
  • fix(cli): rename "docs" skill to "find-docs" in setup output and prompts
• 04130b5: Consolidate skills under /skills with canonical sources: rename docs→find-docs, ctx7-cli→context7-cli, add context7-mcp as canonical MCP skill. MCP setup now downloads skill from GitHub instead of using hardcoded content.
• d418405: Add CLI mode to ctx7 setup for installing the docs skill without MCP configuration

Patch Changes

• 31b4fb8: Align CLI library output format with MCP: use labeled fields (Title, Context7-compatible library ID, Description, Code Snippets, Source Reputation, Benchmark Score, Versions) and categorical reputation labels (High/Medium/Low/Unknown) instead of numeric trust scores
• 9de3f06: Display warning when public library access filter is being used to filter libraries.
• 05a4406: Remove default selection of Universal agent target during skills install prompt
• 9aae852: Show source repository next to skill name in search and suggest results for easier disambiguation

Patch Changes

• 9de3f06: Display warning when public library access filter is being used to filter libraries.

Patch Changes

• df60e3e: Add library and docs commands for querying library documentation from the terminal

Patch Changes

• 9523522: Reject GET requests on MCP endpoints with 405 to eliminate idle SSE connection timeouts
• 59d0327: Include source field in search result response

Patch Changes

• 617d8ed: Remove unnecessary warning and update tool descriptions

Patch Changes

• c66950a: Install documentation-lookup skill during ctx7 setup

Minor Changes

• 3d66191: Add ctx7 setup command for configuring Context7 MCP and rules across Claude Code, Cursor, and OpenCode

Patch Changes

• 4663c15: - Adopt .agents/skills as universal install target, supporting multiple agents with a single installation
  • Replace --codex, --opencode, and --amp flags with single --universal flag
  • Improve checkbox UI with aligned column headers for better readability

Patch Changes

• 0981656: Add skills suggest command that scans your project's dependencies (package.json, requirements.txt, pyproject.toml) and recommends relevant skills. Results show install counts, trust scores, and which dependency each skill matches.

Patch Changes

• 6328ed1: Skill search & generate command improvements:
  • Add "Installs" and "Trust(0-10)" columns to skill search results with aligned column headers
  • Auto-login via OAuth when the generate command requires authentication instead of showing an error
  • Reorder question options so the recommended choice always appears first with a "✓ Recommended" badge
  • Add "View skill" action that opens generated content in the user's default editor ($EDITOR)
  • Revamp generate wizard copy: do/don't examples for skill descriptions, rename "libraries" to "sources", and clarify follow-up question and generation spinner text

Patch Changes

• 2f7cc42: Show exact install counts instead of rounded values, sort skills by install count in the install command, and display "installs" column header inline with the prompt
• 85b905e: Add CLI telemetry for usage metrics collection (commands, searches, installs, generation feedback) via fire-and-forget events to /api/v2/cli/events. Respects CTX7_TELEMETRY_DISABLED env var.

Minor Changes

• 8ba484c: Add AI-powered skill generation with skills generate command, including library search, clarifying questions, real-time query progress, feedback loop, and weekly quota management.
• aacfd31: Add OAuth 2.0 authentication with login, logout, and whoami commands.

Patch Changes

• 572c3ca: Simplify skills list command to show all detected IDE skill directories without prompts.

Patch Changes

• 02148ff: Bump zod from 3.x to 4.x

Patch Changes

• 02148ff: Bump zod from 3.x to 4.x

Minor Changes

• ef82f30: Add OAuth 2.0 authentication support for MCP server
  • Add new /mcp/oauth endpoint requiring JWT authentication
  • Implement JWT validation against authorization server JWKS
  • Add OAuth Protected Resource Metadata endpoint (RFC 9728) at /.well-known/oauth-protected-resource
  • Include WWW-Authenticate header for OAuth discovery

Patch Changes

• 07a53dc: Upgrade ai peer dependency to >=6.0.0 for AI SDK v6 compatibility

Minor Changes

• 9412e62: feat: Change SDK default response type from "txt" to "json" for both searchLibrary and getContext methods. AI SDK tools now explicitly use type: "txt" for LLM-friendly text responses.

Patch Changes

• Updated dependencies [9412e62]
  • @upstash/[email protected]

Minor Changes

• 9412e62: feat: Change SDK default response type from "txt" to "json" for both searchLibrary and getContext methods. AI SDK tools now explicitly use type: "txt" for LLM-friendly text responses.

Minor Changes

• b3cd38a: feat: Rename tools to match MCP naming conventions
  • Rename resolveLibrary to resolveLibraryId with new query parameter
  • Rename getLibraryDocs to queryDocs with new query parameter (replaces topic)
  • Rename RESOLVE_LIBRARY_DESCRIPTION to RESOLVE_LIBRARY_ID_DESCRIPTION
  • Rename GET_LIBRARY_DOCS_DESCRIPTION to QUERY_DOCS_DESCRIPTION
  • Update type re-exports to match new SDK types (Library, Documentation, GetContextOptions)
  • Remove deprecated defaultMaxResults option from Context7ToolsConfig and Context7AgentConfig
  • Add rate limiting guidance to tool descriptions

Patch Changes

• Updated dependencies [b3cd38a]
  • @upstash/[email protected]

Minor Changes

• b3cd38a: feat: Simplify SDK API
  • Replace getDocs() with getContext(query, libraryId, options) - now takes a query parameter for relevance-based retrieval
  • Update searchLibrary(query, libraryName) to take both query and libraryName parameters
  • Replace response types: Library and Documentation instead of SearchResult, CodeDocsResponse, InfoDocsResponse, etc.
  • Remove pagination, mode, topic, and limit options from context retrieval
  • Simplify GetContextOptions to only include type: "json" | "txt"

Patch Changes

• 368b143: Collect client and server version metrics