Skip to content

Context7 Platform -- Up-to-date docs for AI agents

feat: improve documentation clarity

enesgulesd26d5deacbaa
diff --git a/docs/adding-libraries.mdx b/docs/adding-libraries.mdx
index 55814ac8..08d0b5c4 100644
--- a/docs/adding-libraries.mdx
+++ b/docs/adding-libraries.mdx
@@ -48,17 +48,17 @@ Here's an example `context7.json` file with all available options:
## Field Descriptions
-- **`projectTitle`** (string): Suggested display name for your project in Context7. Only used when LLM can not generate a name with high confidence.
+- **`projectTitle`** (string): Suggested display name for your project in Context7. Only used when the LLM cannot generate a name with high confidence.
-- **`description`** (string): Suggested description for your project in Context7. Only used when LLM can not generate a name with high confidence.
+- **`description`** (string): Suggested description for your project in Context7. Only used when the LLM cannot generate a description with high confidence.
- **`branch`** (string): The name of the git branch to parse. If not provided, the default branch will be used.
-- **`folders`** (array): Specific folder paths to include when parsing documentation. If empty, Context7 will scan the entire repository for relevant documentation files. Note: Root-level markdown files are always included regardless of this setting.
+- **`folders`** (array): Specific folder paths to include when parsing. If empty, Context7 scans the entire repository. Root-level markdown files are always included.
-- **`excludeFolders`** (array): Patterns to exclude folders and paths from documentation parsing. Supports simple names, paths, and glob patterns (see Exclusion Patterns section below).
+- **`excludeFolders`** (array): Patterns to exclude from documentation parsing. Supports simple names, paths, and glob patterns (see Exclusion Patterns below).
-- **`excludeFiles`** (array): Specific file names to exclude from documentation parsing. Only include the filename (not the path). Useful for excluding files like `CHANGELOG.md`, license files, or other non-documentation content.
+- **`excludeFiles`** (array): Specific file names to exclude. Use only the filename, not the full path. Examples: `CHANGELOG.md`, license files, or non-documentation content.
- **`rules`** (array): Best practices or important guidelines that coding agents should follow when using your library. These appear as recommendations in the documentation context provided to coding agents.
@@ -122,10 +122,10 @@ These defaults help ensure that coding agents receive relevant, current document
## Best Practices
-1. **Keep descriptions concise**: Aim for one clear sentence that explains your library's purpose to coding agents
-2. **Exclude irrelevant folders**: Use `excludeFolders` to avoid indexing source code, tests, or build artifacts
-3. **Add helpful rules**: Include common gotchas or best practices that coding agents should know when generating code
-4. **Maintain version history**: Keep important previous versions accessible for projects that need older APIs
+1. **Keep descriptions concise**: One sentence explaining your library's purpose. Example: "Serverless Redis SDK for edge and browser environments."
+2. **Exclude irrelevant folders**: Use `excludeFolders` to skip source code, tests, or build artifacts. Example: `["src", "test", "dist", "node_modules"]`
+3. **Add helpful rules**: Include common gotchas or best practices. Example: "Always call `.close()` on Redis connections" or "Use environment variables for API keys, never hardcode them"
+4. **Maintain version history**: Keep important previous versions accessible. Users on v1.x still need v1.x documentation, not just the latest v2.x
## Adding a Version
@@ -151,9 +151,9 @@ To add a new version to your existing library:
Keep documentation helpful for agents:
-- Review the rendered documentation at least once per release.
-- Trim noisy sections (e.g., large changelogs) with `excludeFiles` or `excludeFolders`.
-- Update `rules` when best practices change so AI assistants surface the right guidance.
+- Review the rendered documentation after each release
+- Trim noisy sections (e.g., large changelogs) with `excludeFiles` or `excludeFolders`
+- Update `rules` when best practices change so AI assistants surface current guidance
## Need Help?
diff --git a/docs/api-guide.mdx b/docs/api-guide.mdx
index 524ee61a..83cf5fd9 100644
--- a/docs/api-guide.mdx
+++ b/docs/api-guide.mdx
@@ -14,8 +14,7 @@ Authorization: Bearer CONTEXT7_API_KEY
Get your API key at [context7.com/dashboard](https://context7.com/dashboard).
<Warning>
- Treat your API key like a password. Store it in an environment variable or secret manager and
- rotate it immediately if you suspect it has leaked.
+ Store your API key in an environment variable or secret manager. Rotate it if compromised.
</Warning>
## Rate Limits
@@ -36,17 +35,17 @@ When you exceed rate limits, the API returns a `429` status code with a `retryAf
## Best Practices
-### 1. Use Appropriate Token Limits
+### Use Appropriate Token Limits
Request only the amount of documentation you need to minimize token usage and response times:
```bash
-# Request only 2000 tokens for quick lookups
+# Fetches 2000 tokens of documentation
curl "https://context7.com/api/v1/vercel/next.js?tokens=2000" \
-H "Authorization: Bearer CONTEXT7_API_KEY"
```
-### 2. Specify Topics
+### Specify Topics
Use the `topic` parameter to get more relevant results and reduce unnecessary content:
@@ -56,11 +55,11 @@ curl "https://context7.com/api/v1/vercel/next.js?topic=routing" \
-H "Authorization: Bearer CONTEXT7_API_KEY"
```
-### 3. Cache Responses
+### Cache Responses
Store documentation locally to reduce API calls and improve performance. Documentation updates are relatively infrequent, so caching for several hours or days is usually appropriate.
-### 4. Handle Rate Limits
+### Handle Rate Limits
Implement exponential backoff for rate limit errors:
@@ -82,9 +81,9 @@ def fetch_with_retry(url, headers, max_retries=3):
raise Exception("Max retries exceeded")
```
-### 5. Use Specific Versions
+### Use Specific Versions
-When possible, specify exact versions for consistent results across deployments:
+Specify exact versions for consistent results across deployments:
```bash
# Pin to a specific version
diff --git a/docs/contact.mdx b/docs/contact.mdx
index 3a8693ec..0aaeb6b4 100644
--- a/docs/contact.mdx
+++ b/docs/contact.mdx
@@ -3,13 +3,13 @@ title: Contact & Support
sidebarTitle: Contact
---
-Need help or have questions? We're here to support you.
+Get help and support for Context7.
## Community Support
### Discord
-Join our vibrant community on Discord for:
+Join our Discord community for:
- Quick answers to common questions
- Tips and best practices from other users
- Announcements about new features and updates
@@ -63,10 +63,8 @@ We take all security reports seriously and will respond promptly.
## Feedback
-We value your feedback! Help us improve Context7:
+Help improve Context7:
- Use the thumbs rating on documentation pages
- Suggest edits directly on any page
- Raise issues for bugs or feature requests
-
-Your input helps us build a better product for everyone.
diff --git a/docs/installation.mdx b/docs/installation.mdx
index 7eed1adf..f46b5bea 100644
--- a/docs/installation.mdx
+++ b/docs/installation.mdx
@@ -7,10 +7,10 @@ Get Context7 MCP up and running in your favorite AI coding assistant in minutes.
## Requirements
-Before you begin, make sure you have:
+Before you begin:
- **MCP Client**: Cursor, Claude Code, VS Code, Windsurf, or another MCP-compatible IDE
-- **API Key** (Recommended): For higher rate limits and private repositories. Get yours at [context7.com/dashboard](https://context7.com/dashboard)
+- **API Key (Recommended)**: For higher rate limits and private repositories. Get yours at [context7.com/dashboard](https://context7.com/dashboard)
<Warning>
**SSE Protocol Removed**
@@ -29,7 +29,7 @@ Install Context7 MCP automatically for any client via [Smithery](https://smither
npx -y @smithery/cli@latest install @upstash/context7-mcp --client <CLIENT_NAME> --key <YOUR_SMITHERY_KEY>
```
-Find your Smithery key at [smithery.ai](https://smithery.ai/server/@upstash/context7-mcp).
+Get your Smithery key from your account at [smithery.ai](https://smithery.ai/server/@upstash/context7-mcp).
## Client-Specific Installation
@@ -37,7 +37,7 @@ Find your Smithery key at [smithery.ai](https://smithery.ai/server/@upstash/cont
<Accordion title="Cursor" icon="code">
Go to: `Settings` → `Cursor Settings` → `MCP` → `Add new global MCP server`
- Since Cursor 1.0, you can use one-click installation or manually add to `~/.cursor/mcp.json`.
+ In Cursor 1.0 and later, use the one-click button below or manually edit `~/.cursor/mcp.json`.
**Remote Server (Recommended)**:
diff --git a/docs/overview.mdx b/docs/overview.mdx
index f3f7f13f..7e7415f0 100644
--- a/docs/overview.mdx
+++ b/docs/overview.mdx
@@ -7,7 +7,7 @@ Context7 brings up-to-date, version-specific documentation and code examples dir
## ❌ Without Context7
-LLMs rely on outdated or generic information about the libraries you use. You get:
+LLMs rely on outdated or generic information about the libraries you use:
- ❌ Code examples are outdated and based on year-old training data
- ❌ Hallucinated APIs that don't even exist
@@ -30,7 +30,7 @@ Configure a Cloudflare Worker script to cache JSON API responses for five minute
Context7 fetches up-to-date code examples and documentation right into your LLM's context.
- 1️⃣ Write your prompt naturally
-- 2️⃣ Tell the LLM to `use context7`
-- 3️⃣ Get working code answers
+- 2️⃣ Add `use context7` to your prompt
+- 3️⃣ Get working code with current APIs
No tab-switching, no hallucinated APIs that don't exist, no outdated code generation.
diff --git a/docs/plans-pricing.mdx b/docs/plans-pricing.mdx
index 59cef6dd..f7bc15b4 100644
--- a/docs/plans-pricing.mdx
+++ b/docs/plans-pricing.mdx
@@ -75,17 +75,20 @@ Pick the plan that matches how you work today and scale when your needs grow.
<AccordionGroup>
<Accordion title="How do you keep the privacy of my queries?" icon="shield">
- No user queries are sent to our servers. Instead, only generic topics are transmitted. Your actual prompts and code remain private on your local machine.
+ The MCP client sends only generic topics to our servers, not your actual queries. Your prompts and code remain private on your local machine.
</Accordion>
<Accordion title="How do you ensure the code snippets are safe?" icon="check">
- We follow several processes to ensure safety: - Run an LLM-based model to detect injection
- attempts before storing code - Assign a trustScore to repository authors - Allow users to filter
- trusted libraries by organization and repository - Regular security audits and monitoring
+ We follow several processes to ensure safety:
+
+ - Run an LLM-based model to detect injection attempts before storing code
+ - Assign a trustScore to repository authors
+ - Allow users to filter trusted libraries by organization and repository
+ - Regular security audits and monitoring
</Accordion>
<Accordion title="What are your security practices?" icon="lock">
- We are part of Upstash Inc., a four-year-old Silicon Valley company. You can review our certifications and practices on our [Trust page](https://trust.upstash.com). We follow SOC-2 compliance standards and industry best practices.
+ We are part of Upstash Inc., based in Silicon Valley since 2020. Review our certifications and practices on our [Trust page](https://trust.upstash.com). We follow SOC-2 compliance standards and industry best practices.
</Accordion>
</AccordionGroup>
diff --git a/docs/security.mdx b/docs/security.mdx
index 34f6d993..eb4ee438 100644
--- a/docs/security.mdx
+++ b/docs/security.mdx
@@ -148,10 +148,10 @@ For European users, Context7 provides:
## Reporting Security Issues
-If you discover a security vulnerability, please report it responsibly:
+If you discover a security vulnerability:
1. **Do not** publicly disclose the issue
-2. Email security details to: [GitHub Security](https://github.com/upstash/context7/security)
+2. Report via: [GitHub Security](https://github.com/upstash/context7/security)
3. Include detailed steps to reproduce the issue
4. Allow reasonable time for us to address the issue
diff --git a/docs/troubleshooting.mdx b/docs/troubleshooting.mdx
index b7f031bd..80c6505e 100644
--- a/docs/troubleshooting.mdx
+++ b/docs/troubleshooting.mdx
@@ -29,11 +29,11 @@ If you encounter `ERR_MODULE_NOT_FOUND`, try using `bunx` instead of `npx`:
}
```
-This often resolves module resolution issues in environments where `npx` doesn't properly install or resolve packages.
+Use this when you see module resolution errors with npx.
### ESM Resolution Issues
-For errors like `Error: Cannot find module 'uriTemplate.js'`, try the `--experimental-vm-modules` flag:
+For errors such as `Error: Cannot find module 'uriTemplate.js'`, try the `--experimental-vm-modules` flag:
```json
{
@@ -61,9 +61,7 @@ Use the `--experimental-fetch` flag to bypass TLS-related problems:
}
```
-## General Solutions
-
-### Quick Fixes
+## Quick Fixes
1. **Try adding `@latest`** to the package name to ensure you're using the most recent version:
@@ -326,11 +324,11 @@ Test the remote server is reachable:
curl https://mcp.context7.com/mcp/ping
```
-Should return: `{"status": "ok", "message": "pong"}`
+Expected response: `{"status": "ok", "message": "pong"}`
-## Still Having Issues?
+## Additional Support
-If none of these solutions work:
+If these solutions don't resolve your issue:
1. **Check GitHub Issues**: Search for similar problems at [github.com/upstash/context7/issues](https://github.com/upstash/context7/issues)
2. **Create a new issue**: Include your:
diff --git a/docs/usage.mdx b/docs/usage.mdx
index 8d18b483..73336d3a 100644
--- a/docs/usage.mdx
+++ b/docs/usage.mdx
@@ -2,19 +2,19 @@
title: How-To Guides
---
-Learn practical techniques to maximize your Context7 experience.
+Practical techniques for using Context7 effectively.
## Quick Wins
- [Auto-invoke Context7](#auto-invoke-context7) so you never forget to pull fresh docs
- [Target the exact library ID](#use-specific-library-ids) to skip search steps
- [Tune networking and proxies](#configure-https-proxy) for locked-down environments
-- [Control token usage](#optimize-token-usage) to keep responses snappy
+- [Control token usage](#control-token-usage) to keep responses snappy
- [Bring in private repositories](#work-with-private-repositories) for internal docs
## Auto-Invoke Context7
-Stop adding "use context7" to every prompt by setting up an automatic rule.
+Stop adding `use context7` to every prompt by setting up an automatic rule.
### Add a Rule to Your MCP Client
@@ -50,8 +50,7 @@ Stop adding "use context7" to every prompt by setting up an automatic rule.
</AccordionGroup>
<Tip>
- Customize the rule to match your workflow. For example, only auto-invoke for specific libraries or
- frameworks.
+ Customize the rule for your workflow. Examples: "Auto-invoke only for Next.js questions" or "Use context7 when the prompt mentions 'API' or 'documentation'."
</Tip>
## Use Specific Library IDs
@@ -153,9 +152,9 @@ Both lowercase and uppercase environment variables are supported.
connectivity before restarting your IDE.
</Tip>
-## Optimize Token Usage
+## Control Token Usage
-Control documentation size to manage costs and response times.
+Manage documentation size to control costs and response times.
### Set Token Limits in Dashboard
@@ -182,7 +181,7 @@ curl "https://context7.com/api/v1/vercel/next.js?tokens=3000&topic=routing" \
-H "Authorization: Bearer YOUR_API_KEY"
```
-Lower token counts = faster responses and lower costs.
+Lower token counts result in faster responses and lower costs.
## Work with Private Repositories
@@ -309,28 +308,28 @@ Visit the [Dashboard](https://context7.com/dashboard) to monitor your usage and
### Security
-- Never commit API keys to version control
-- Use environment variables for keys
-- Rotate keys regularly
-- Give team members minimum required permissions
+- Never commit API keys to version control - Use `.env` files and add them to `.gitignore`
+- Use environment variables for keys - `export CONTEXT7_API_KEY=your_key_here`
+- Rotate keys regularly - Rotate every 90 days or immediately if compromised
+- Give team members minimum required permissions - Use "Developer" role for non-admins
See [Security](/security) for comprehensive security guidelines.
### Performance
-- Use specific library IDs when possible
-- Set appropriate token limits
-- Use topic filtering to focus results
-- Cache frequently-used documentation locally
+- Use specific library IDs when possible - `/vercel/next.js` instead of "next.js docs"
+- Set appropriate token limits - small amounts for quick lookups, higher amounts for comprehensive docs
+- Use topic filtering to focus results - Add specific topics like "authentication" or "routing"
+- Cache frequently-used documentation locally - Cache for 6-24 hours to reduce API calls
### Library Management
-- Keep `context7.json` up to date in your repositories
-- Exclude unnecessary folders (tests, build artifacts)
-- Add helpful rules for AI agents
-- Maintain version tags for important releases
+- Keep `context7.json` up to date in your repositories - Update when adding new docs or changing structure
+- Exclude unnecessary folders (tests, build artifacts) - `excludeFolders: ["test", "dist", "node_modules"]`
+- Add helpful rules for AI agents - Example: "Always validate user input" or "Use TypeScript strict mode"
+- Maintain version tags for important releases - Keep v1.x docs available even after releasing v2.x
-See [Library Owners](/library-owners) for configuration details.
+See [Library Owners](/adding-libraries) for configuration details.
## Need Help?