Add new mastra agents #29
Conversation
WalkthroughThis change introduces three new Agentuity-based AI agent projects: Startup News Scraper, Blog Publisher, and Cat Fact Agent. Each project includes configuration files, documentation, code for agent logic, and workflow or tool integrations. The updates also provide detailed SDK usage guidelines, editor settings, and project-specific TypeScript configurations. Changes
Sequence Diagram(s)Startup News Scraper Agent FlowsequenceDiagram
participant User
participant n8n Workflow
participant StartupScraperAgent
participant OpenAI
User->>n8n Workflow: Triggers daily schedule
n8n Workflow->>StartupScraperAgent: HTTP POST (optionally with perFeed)
StartupScraperAgent->>StartupScraperAgent: Fetch and filter RSS feeds
StartupScraperAgent->>OpenAI: Summarize startup news articles
OpenAI-->>StartupScraperAgent: Summaries
StartupScraperAgent-->>n8n Workflow: JSON response with summaries
n8n Workflow->>n8n Workflow: Format HTML digest
n8n Workflow->>Slack: Send digest
n8n Workflow->>Email: Send digest
n8n Workflow->>Discord: Send digest
Blog Publisher Multi-Agent WorkflowsequenceDiagram
participant User
participant PublisherAgent
participant CopywriterAgent
participant EditorAgent
User->>PublisherAgent: Provide blog topic
PublisherAgent->>CopywriterAgent: Generate draft copy
CopywriterAgent-->>PublisherAgent: Draft copy
PublisherAgent->>EditorAgent: Edit draft copy
EditorAgent-->>PublisherAgent: Edited copy
PublisherAgent-->>User: Return final blog post
Cat Fact Agent FlowsequenceDiagram
participant User
participant CatFactAgent
participant CatFactTool
participant OpenAI
User->>CatFactAgent: Ask for cat fact or info
CatFactAgent->>CatFactTool: Fetch real cat fact
CatFactTool-->>CatFactAgent: Cat fact
CatFactAgent->>OpenAI: Compose response with fact
OpenAI-->>CatFactAgent: Response
CatFactAgent-->>User: Return answer with cat fact
Suggested reviewers
Poem
✨ Finishing Touches
🧪 Generate Unit Tests
🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 14
🔭 Outside diff range comments (2)
frameworks/mastra/cat-fact-agent/.cursor/rules/agent.mdc (1)
37-37: Fix missing closing markdown fence.The document is missing a closing triple backtick (```) which will cause rendering issues.
Apply this diff to fix the markdown syntax:
The AgentContext has information specific to the incoming Agent request and a set of helper methods for accessing AI services like KeyValue storage and Vector storage. +```agents/Startup_News_Scraper/.cursor/rules/sdk.mdc (1)
98-98: Fix missing closing markdown fence.The document is missing a closing triple backtick (```) at the end which will cause rendering issues.
Apply this diff to fix the markdown syntax:
For complete documentation, visit: https://agentuity.dev/SDKs/javascript/api-reference +```
♻️ Duplicate comments (1)
frameworks/mastra/blog-publisher/package.json (1)
4-4: Inconsistent main and module entry points.Same issue as in the cat-fact-agent: the
mainfield points toindex.jswhile themodulefield points toindex.ts.Apply this diff to align the entry points:
- "main": "index.js", + "main": "index.ts",Also applies to: 36-36
🧹 Nitpick comments (30)
frameworks/mastra/blog-publisher/agentuity.yaml (2)
8-15: Add a project description.An empty description (
description: "") reduces discoverability and context for new contributors. Consider filling in a brief summary of the blog-publisher project's purpose.
71-76: Include agent metadata fields.Consider adding a
description(and any relevant tags or version) under theagentsblock to document the supervisor agent’s role.agents/Startup_News_Scraper/biome.json (1)
24-27: Expand ignored files list.Currently only
.agentuity/**is ignored. It’s best practice to also ignore tool directories (e.g.,.cursor/**) and common folders likenode_modules/**anddist/**to prevent linting/formatting churn.frameworks/mastra/blog-publisher/.editorconfig (1)
6-12: Harden editor settings.
trim_trailing_whitespaceisfalse; enabling it helps avoid accidental whitespace diffs.- Setting
insert_final_newline = trueensures POSIX-compliant files.- Since
indent_style = tab, you may omit or setindent_size = tabto avoid confusion.agents/Startup_News_Scraper/tsconfig.json (1)
1-24: Refine TS compiler configuration.
- Add an explicit
"include": ["src"]to limit compilation scope.- Extend the
"exclude"array to ignore the.cursordirectory.- Confirm whether
"jsx": "react-jsx"is needed—remove or adjust it if you’re not using React.frameworks/mastra/blog-publisher/src/agents/supervisor/agents.ts (1)
1-74: LGTM! Well-architected multi-agent system.This implementation demonstrates excellent use of the supervisor pattern with proper separation of concerns. The use of different AI models for different roles (Claude for creative tasks, GPT-4o-mini for editing) is a smart design choice. Schema validation with Zod ensures type safety throughout the workflow.
Consider these optional improvements:
- Replace console.log with structured logging:
- console.log("🛠️ Copywriter tool executing..."); + ctx.logger?.info("🛠️ Copywriter tool executing...");
- Add error handling in tool execution:
execute: async ({ context }) => { + try { console.log("🛠️ Copywriter tool executing..."); // ... existing code return { copy: result.text }; + } catch (error) { + throw new Error(`Copywriter tool failed: ${error.message}`); + } },frameworks/mastra/cat-fact-agent/index.ts (1)
3-9: Consider moving global interface extension to a type definition file.The global
NodeJS.Processinterface extension would be better placed in a dedicated type definition file (e.g.,types/global.d.ts) for better organization and reusability.Create a new file
types/global.d.ts:declare global { namespace NodeJS { interface Process { isBun: boolean; } } } export {};Then remove the global declaration from this file and add the types directory to your
tsconfig.json.agents/Startup_News_Scraper/src/agents/startup_scraper/types.ts (1)
8-16: Consider consistent nullable handling in StartupSummary interface.The interface mixes optional properties (
industry?) with explicit nullable types (summary: string | null). For consistency, consider using either optional properties or explicit nullable types throughout.export interface StartupSummary { name: string; industry?: string; - summary: string | null; + summary?: string; founders?: string | null; why_it_matters?: string | null; funding_source?: string | null; source: string; }Or alternatively, make all optional fields explicitly nullable:
export interface StartupSummary { name: string; - industry?: string; + industry: string | null; - summary: string | null; + summary: string | null; - founders?: string | null; + founders: string | null; - why_it_matters?: string | null; + why_it_matters: string | null; - funding_source?: string | null; + funding_source: string | null; source: string; }agents/Startup_News_Scraper/index.ts (1)
3-9: Consider moving global type declaration to a separate declaration file.The global type augmentation for NodeJS.Process interface is better placed in a
global.d.tsortypes.d.tsfile for better organization and reusability.Create a
global.d.tsfile:declare global { namespace NodeJS { interface Process { isBun: boolean; } } } export {};frameworks/mastra/cat-fact-agent/README.md (1)
13-13: Fix missing comma for better readability.There's a missing comma that improves sentence flow and clarity.
-It demonstrates how a system prompt can shape an agent's personality and behavior in this case, turning the agent into a **verified cat expert** that always provides fun feline facts in its responses. +It demonstrates how a system prompt can shape an agent's personality and behavior, in this case turning the agent into a **verified cat expert** that always provides fun feline facts in its responses.frameworks/mastra/cat-fact-agent/src/agents/cat-agent/cat.ts (1)
12-19: Consider adding input validation schema even for parameterless tools.While the tool doesn't require inputs, using a more explicit schema can improve clarity and future extensibility.
const catFact = createTool({ id: "Get cat facts", - inputSchema: z.object({}), // no inputs + inputSchema: z.object({}), // No inputs required description: "Fetches cat facts", execute: async () => ({ catFact: await getCatFact(), }), });agents/Startup_News_Scraper/src/agents/startup_scraper/fetchFeeds.ts (1)
6-12: Consider making the keyword list configurable.The hardcoded
KEEPkeyword array could benefit from being configurable, especially for different deployment environments or user preferences.-const KEEP = [ - 'startup', 'startups', - 'ai', 'artificial intelligence', - 'tech', 'technology', - 'vc', 'venture capital', -]; +const KEEP = process.env.STARTUP_KEYWORDS?.split(',') || [ + 'startup', 'startups', + 'ai', 'artificial intelligence', + 'tech', 'technology', + 'vc', 'venture capital', +];agents/Startup_News_Scraper/n8n_workflow/workflow.json (1)
62-62: Consider extracting JavaScript code to a separate file.The inline JavaScript code for HTML formatting is quite complex and would benefit from being extracted to a separate, testable file.
The current inline code could be moved to a module like:
// formatStartupDigest.js function formatStartupDigest(startups) { let html = ` <h2>🚀 Daily Startup Digest</h2> <p>Here are today's top startup highlights:</p> `; for (const s of startups) { html += ` <hr> <h3>${s.name}</h3> <p><strong>Industry:</strong> ${s.industry || 'N/A'}</p> <p><strong>Summary:</strong><br>${s.summary}</p> ${s.why_it_matters ? `<p><strong>Why it matters:</strong><br>${s.why_it_matters}</p>` : ''} ${s.founders ? `<p><strong>Founders:</strong> ${s.founders}</p>` : ''} <p><strong>Source:</strong> <a href="${s.source}" target="_blank">${s.source}</a></p> `; } return html; }frameworks/mastra/blog-publisher/src/agents/supervisor/index.ts (1)
1-1: Incorrect file path comment.The comment indicates this is
src/index.tsbut the actual file path issrc/agents/supervisor/index.ts.-// src/index.ts +// src/agents/supervisor/index.tsframeworks/mastra/blog-publisher/.cursor/rules/agent.mdc (1)
26-36: Interface descriptions could be more detailed.The descriptions of AgentRequest, AgentResponse, and AgentContext are quite brief. Consider adding more specific details about available methods and common use cases.
Consider expanding the descriptions:
### AgentRequest -The AgentRequest interface provides a set of helper methods and public variables which can be used for working with data has been passed to the Agent. +The AgentRequest interface provides helper methods for accessing request data: +- `data.text()` - Extract text content from the request +- `data.json()` - Parse JSON from the request body +- `headers` - Access HTTP headers +- `query` - Access query parameters ### AgentResponse -The AgentResponse interface provides a set of helper methods for responding with different data formats from the Agent. +The AgentResponse interface provides methods for sending responses: +- `resp.text(string)` - Send plain text response +- `resp.json(object)` - Send JSON response +- `resp.html(string)` - Send HTML response ### AgentContext -The AgentContext has information specific to the incoming Agent request and a set of helper methods for accessing AI services like KeyValue storage and Vector storage. +The AgentContext provides request-specific information and services: +- `ctx.logger` - Structured logging interface +- `ctx.kv` - Key-value storage access +- `ctx.vector` - Vector database operations +- Request metadata and environment informationagents/Startup_News_Scraper/.cursor/rules/agent.mdc (1)
26-36: Add more specific documentation for the interfaces.While the interface descriptions are helpful, they could be more detailed. Consider adding specific examples of common use cases for each interface to help developers understand practical applications.
### AgentRequest -The AgentRequest interface provides a set of helper methods and public variables which can be used for working with data has been passed to the Agent. +The AgentRequest interface provides helper methods and properties for accessing request data, including JSON/text parsing, metadata extraction, and trigger information. ### AgentResponse -The AgentResponse interface provides a set of helper methods for responding with different data formats from the Agent. +The AgentResponse interface provides methods for creating various response types (JSON, text, HTML, binary) and agent handoffs. ### AgentContext -The AgentContext has information specific to the incoming Agent request and a set of helper methods for accessing AI services like KeyValue storage and Vector storage. +The AgentContext provides access to logging, key-value storage, vector storage, remote agents, and tracing capabilities for the current request.agents/Startup_News_Scraper/src/agents/startup_scraper/index.ts (2)
12-15: Consider adding return type annotation.The
welcomefunction should have an explicit return type for better type safety and documentation.-export const welcome = () => ({ +export const welcome = (): { welcome: string } => ({ welcome: 'POST {} (or {"perFeed":3}) to receive the three freshest startup articles from each feed, summarized.', });
27-32: Add input validation for perFeed parameter.Consider adding validation to ensure
perFeedis a reasonable positive integer to prevent potential issues downstream.const perFeed = body.perFeed ?? 3; // ← new default + + // Validate perFeed parameter + if (typeof perFeed !== 'number' || perFeed < 1 || perFeed > 50) { + ctx.logger.warn('Invalid perFeed value %s, using default of 3', perFeed); + perFeed = 3; + }agents/Startup_News_Scraper/README.md (5)
70-70: Add missing comma for better readability.There's a missing comma that makes the sentence harder to parse.
-1. In your **Agentuity** dashboard open the deployed agent, click the **➕** icon under *IO Visualization* and choose **API**. +1. In your **Agentuity** dashboard, open the deployed agent, click the **➕** icon under *IO Visualization* and choose **API**.
83-84: Fix duplicated word in JSON description.The word "JSON" appears twice in the description.
- - Body Content Type – select **JSON** - - JSON Body – `{}` (a single empty JSON object) + - Body Content Type – select **JSON** + - Body – `{}` (a single empty JSON object)
87-89: Fix section heading repetition.The phrase "Code Node" is repeated unnecessarily in the heading and first line.
-### 6. Do not adjust Code Node -1. Code node is preconfigured with the logic to transform the JSON formatted summaries to human-readable HTML. +### 6. Do not adjust Code Node +1. This node is preconfigured with the logic to transform the JSON formatted summaries to human-readable HTML.
91-91: Consider hyphenating compound adjective."Third Party" should be hyphenated when used as a compound adjective.
-### 7. Third Party Configuration (Email, Slack, Discord, etc.) +### 7. Third-Party Configuration (Email, Slack, Discord, etc.)
51-56: Verify package manager consistency.The setup instructions use
buncommands, but this might not be available in all environments. Consider providing alternative commands for npm/yarn users.```bash git clone https://github.com/agentuity/examples cd agents/Startup_News_Scraper bun install bun dev +# Or using npm: +# npm install +# npm run dev</blockquote></details> <details> <summary>frameworks/mastra/blog-publisher/.cursor/rules/sdk.mdc (2)</summary><blockquote> `1-5`: **Consider enabling alwaysApply for consistent SDK reference.** The `alwaysApply: false` setting might limit the usefulness of this SDK reference. Consider enabling it for consistent access to SDK documentation. ```diff --- description: Agentuity JavaScript SDK API Reference globs: "src/agents/**/*.ts" -alwaysApply: false +alwaysApply: true ---
17-23: Add return type information to AgentHandler.The
AgentResponseTypeis referenced but not defined. Consider adding more context about what this type represents.```typescript type AgentHandler = ( request: AgentRequest, response: AgentResponse, context: AgentContext -) => Promise<AgentResponseType>; +) => Promise<AgentResponseType>; // AgentResponseType represents the response from any response method</blockquote></details> <details> <summary>agents/Startup_News_Scraper/src/agents/startup_scraper/summarize.ts (3)</summary><blockquote> `19-24`: **Consider logging failed summaries for debugging.** While using `Promise.allSettled` is correct for handling partial failures, consider logging rejected promises to aid debugging. ```diff for (const group of chunks) { const settled = await Promise.allSettled(group.map(callOpenAI)); for (const res of settled) { - if (res.status === 'fulfilled') summaries.push(res.value); + if (res.status === 'fulfilled') { + summaries.push(res.value); + } else { + console.warn('Failed to summarize startup:', res.reason); + } } }
44-51: Consider adding retry logic and rate limiting.OpenAI API calls can fail due to rate limits or temporary issues. Consider implementing exponential backoff retry logic.
You may want to add retry logic and potentially adjust the temperature and token limits based on your needs:
+const MAX_RETRIES = 3; +const RETRY_DELAY = 1000; // 1 second base delay +async function callOpenAIWithRetry(link: StartupLink, retries = 0): Promise<StartupSummary> { + try { const res = await openai.chat.completions.create({ model: 'gpt-4o-mini', messages: [{ role: 'user', content: prompt }], response_format: { type: 'json_object' }, - max_tokens: 400, - temperature: 0.4, + max_tokens: 500, // Slightly higher for comprehensive summaries + temperature: 0.2, // Lower for more consistent output }); + } catch (error) { + if (retries < MAX_RETRIES && (error.status === 429 || error.status >= 500)) { + await new Promise(resolve => setTimeout(resolve, RETRY_DELAY * Math.pow(2, retries))); + return callOpenAIWithRetry(link, retries + 1); + } + throw error; + }
76-84: Add return type annotation and improve regex patterns.The function should have explicit return type annotation, and consider making regex patterns more precise.
-function normaliseIndustry(raw?: string): 'AI' | 'VC' | 'Tech' | 'Startups' | undefined { +function normaliseIndustry(raw?: string): 'AI' | 'VC' | 'Tech' | 'Startups' { if (!raw) return undefined; const t = raw.toLowerCase(); - if (/(^| )(ai|artificial intelligence|machine learning|ml)( |$)/.test(t)) { return 'AI';} - if (/(vc|venture|fund|capital|investor)/.test(t)) { return 'VC';} - if (/(startup|accelerator|incubator|founder program)/.test(t)) { return 'Startups';} - if (/(tech|technology|robot|drone|fintech|crypto|software|hardware|cloud|SaaS)/.test(t)) { return 'Tech';} - return undefined; + if (/\b(ai|artificial intelligence|machine learning|ml|llm|gpt)\b/.test(t)) return 'AI'; + if (/\b(vc|venture|fund|capital|investor|investment)\b/.test(t)) return 'VC'; + if (/\b(startup|accelerator|incubator|founder program|entrepreneurship)\b/.test(t)) return 'Startups'; + if (/\b(tech|technology|robot|drone|fintech|crypto|software|hardware|cloud|saas|platform)\b/i.test(t)) return 'Tech'; + return 'Tech'; // Default fallback instead of undefined }frameworks/mastra/cat-fact-agent/.cursor/rules/sdk.mdc (2)
17-23: Consider adding parameter descriptions to AgentHandler type.The type signature is correct, but adding brief parameter descriptions would improve clarity.
type AgentHandler = ( - request: AgentRequest, - response: AgentResponse, - context: AgentContext + request: AgentRequest, // Incoming request data and metadata + response: AgentResponse, // Response builder methods + context: AgentContext // Runtime context (logging, storage, etc.) ) => Promise<AgentResponseType>;
88-96: Enhance best practices with specific examples.The best practices section could benefit from concrete examples and more detailed guidance.
Consider expanding the best practices section:
## Best Practices - Use TypeScript for better type safety and IDE support - Import types from `@agentuity/sdk` - Use structured error handling with try/catch blocks - Leverage the provided logger for consistent logging - Use the storage APIs for persisting data - Consider agent communication for complex workflows + +### Example Usage + +```typescript +import type { AgentHandler } from '@agentuity/sdk'; + +const handler: AgentHandler = async (request, response, context) => { + try { + const data = await request.data.json(); + context.logger.info('Processing request', { data }); + + // Your agent logic here + + return response.json({ success: true }); + } catch (error) { + context.logger.error('Agent error', error); + return response.json({ error: 'Internal error' }, { status: 500 }); + } +}; +```
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
⛔ Files ignored due to path filters (3)
agents/Startup_News_Scraper/bun.lockis excluded by!**/*.lockframeworks/mastra/blog-publisher/bun.lockis excluded by!**/*.lockframeworks/mastra/cat-fact-agent/bun.lockis excluded by!**/*.lock
📒 Files selected for processing (42)
agents/Startup_News_Scraper/.cursor/rules/agent.mdc(1 hunks)agents/Startup_News_Scraper/.cursor/rules/agentuity.mdc(1 hunks)agents/Startup_News_Scraper/.cursor/rules/sdk.mdc(1 hunks)agents/Startup_News_Scraper/.editorconfig(1 hunks)agents/Startup_News_Scraper/.gitignore(1 hunks)agents/Startup_News_Scraper/README.md(1 hunks)agents/Startup_News_Scraper/agentuity.yaml(1 hunks)agents/Startup_News_Scraper/biome.json(1 hunks)agents/Startup_News_Scraper/index.ts(1 hunks)agents/Startup_News_Scraper/n8n_workflow/workflow.json(1 hunks)agents/Startup_News_Scraper/package.json(1 hunks)agents/Startup_News_Scraper/src/agents/startup_scraper/fetchFeeds.ts(1 hunks)agents/Startup_News_Scraper/src/agents/startup_scraper/index.ts(1 hunks)agents/Startup_News_Scraper/src/agents/startup_scraper/summarize.ts(1 hunks)agents/Startup_News_Scraper/src/agents/startup_scraper/types.ts(1 hunks)agents/Startup_News_Scraper/tsconfig.json(1 hunks)frameworks/mastra/blog-publisher/.cursor/rules/agent.mdc(1 hunks)frameworks/mastra/blog-publisher/.cursor/rules/agentuity.mdc(1 hunks)frameworks/mastra/blog-publisher/.cursor/rules/sdk.mdc(1 hunks)frameworks/mastra/blog-publisher/.editorconfig(1 hunks)frameworks/mastra/blog-publisher/.gitignore(1 hunks)frameworks/mastra/blog-publisher/README.md(1 hunks)frameworks/mastra/blog-publisher/agentuity.yaml(1 hunks)frameworks/mastra/blog-publisher/biome.json(1 hunks)frameworks/mastra/blog-publisher/index.ts(1 hunks)frameworks/mastra/blog-publisher/package.json(1 hunks)frameworks/mastra/blog-publisher/src/agents/supervisor/agents.ts(1 hunks)frameworks/mastra/blog-publisher/src/agents/supervisor/index.ts(1 hunks)frameworks/mastra/blog-publisher/tsconfig.json(1 hunks)frameworks/mastra/cat-fact-agent/.cursor/rules/agent.mdc(1 hunks)frameworks/mastra/cat-fact-agent/.cursor/rules/agentuity.mdc(1 hunks)frameworks/mastra/cat-fact-agent/.cursor/rules/sdk.mdc(1 hunks)frameworks/mastra/cat-fact-agent/.editorconfig(1 hunks)frameworks/mastra/cat-fact-agent/.gitignore(1 hunks)frameworks/mastra/cat-fact-agent/README.md(1 hunks)frameworks/mastra/cat-fact-agent/agentuity.yaml(1 hunks)frameworks/mastra/cat-fact-agent/biome.json(1 hunks)frameworks/mastra/cat-fact-agent/index.ts(1 hunks)frameworks/mastra/cat-fact-agent/package.json(1 hunks)frameworks/mastra/cat-fact-agent/src/agents/cat-agent/cat.ts(1 hunks)frameworks/mastra/cat-fact-agent/src/agents/cat-agent/index.ts(1 hunks)frameworks/mastra/cat-fact-agent/tsconfig.json(1 hunks)
🧰 Additional context used
🪛 LanguageTool
frameworks/mastra/cat-fact-agent/README.md
[uncategorized] ~13-~13: Possible missing comma found.
Context: ...pt can shape an agent's personality and behavior in this case, turning the agent into a ...
(AI_HYDRA_LEO_MISSING_COMMA)
agents/Startup_News_Scraper/README.md
[uncategorized] ~70-~70: A comma might be missing here.
Context: ...y API endpoint 1. In your Agentuity dashboard open the deployed agent, click the *➕...
(AI_EN_LECTOR_MISSING_PUNCTUATION_COMMA)
[duplication] ~83-~83: Possible typo: you repeated a word.
Context: ...on** - Body Content Type – select JSON - JSON Body – {} (a single empty JSON objec...
(ENGLISH_WORD_REPEAT_RULE)
[grammar] ~87-~87: This phrase is duplicated. You should probably use “Code Node” only once.
Context: ...– leave blank ### 6. Do not adjust Code Node 1. Code node is preconfigured with the logic to tran...
(PHRASE_REPETITION)
[uncategorized] ~91-~91: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...eadable HTML. 2. Do not change. ### 7. Third Party Configuration (Email, Slack, Discord, e...
(EN_COMPOUND_ADJECTIVE_INTERNAL)
🔇 Additional comments (54)
frameworks/mastra/cat-fact-agent/.cursor/rules/agentuity.mdc (2)
1-5: Front matter is correctly configured. The YAML header defines the description, file globs, and enforcement flag consistently with other Agentuity rule files.
7-9: Directive is clear. The notice not to suggest edits is appropriate for generated config files and prevents accidental modifications.agents/Startup_News_Scraper/.editorconfig (1)
1-12: Consistent EditorConfig settings. Defining this as the root ensures tab-based indentation (size 2), LF line endings, UTF-8 charset, and controlled whitespace handling—matching the style of other agent projects.frameworks/mastra/blog-publisher/biome.json (1)
1-27: Biome configuration aligns with standards. The schema version, import organization, linting rules, formatter settings (2-space indent, single quotes, semicolons), and ignored paths mirror other agent frameworks.frameworks/mastra/cat-fact-agent/.editorconfig (1)
1-12: EditorConfig matches project conventions. Root declaration, tab indentation, LF endings, and UTF-8 encoding ensure consistency across this agent’s files.frameworks/mastra/cat-fact-agent/biome.json (1)
1-27: Biome setup is correct. The JSON config enables import organization, recommended linting, formatting rules, and excludes.agentuityfiles, consistent with other agent projects.frameworks/mastra/blog-publisher/.cursor/rules/agentuity.mdc (1)
1-9: Guidelines schema is correct
The front-matter metadata and instructional text clearly document theagentuity.yamlconfiguration. No edits needed.agents/Startup_News_Scraper/.gitignore (1)
1-40: .gitignore covers all transient and build artifacts
Patterns for dependencies, outputs, coverage, logs, environment files, caches, IDE settings, macOS metadata, and Agentuity files are comprehensive and consistent.frameworks/mastra/cat-fact-agent/.gitignore (1)
1-40: .gitignore is thorough and consistent
Matches the structure used in other agent projects; no gaps detected.frameworks/mastra/blog-publisher/tsconfig.json (1)
1-24: tsconfig aligns with project conventions
Strict settings and module options match other Agentuity projects. Verify that your TypeScript toolchain is v5.0+ to supportallowImportingTsExtensionsandverbatimModuleSyntax.frameworks/mastra/blog-publisher/.gitignore (1)
1-40: .gitignore is well-structured and consistent
All expected directories and file types (node_modules, build outputs, coverage, env files, caches, IDE/OS files, Agentuity metadata) are properly excluded.frameworks/mastra/blog-publisher/agentuity.yaml (1)
50-57: Verify CPU resource unit.You’ve specified
cpu: 500M, but most Kubernetes-style configs expect millicores as500m(lowercase m). Please confirm with the Agentuity schema or change to500mto avoid unintended resource limits.agents/Startup_News_Scraper/.cursor/rules/agentuity.mdc (1)
1-10: Configuration rule header looks good.The frontmatter and guidance correctly target
agentuity.yamland enforce the rule. No changes needed.frameworks/mastra/cat-fact-agent/tsconfig.json (1)
1-24: LGTM! Well-configured TypeScript setup.The configuration follows modern TypeScript best practices with strict type checking, ESNext targeting, and appropriate bundler module resolution. The inclusion of
@agentuity/sdktypes and exclusion of build directories are properly configured.agents/Startup_News_Scraper/package.json (1)
1-35: LGTM! Well-structured package configuration.The package.json follows best practices with appropriate scripts, dependencies, and tooling setup. The use of Biome for formatting/linting and Bun runtime is modern and efficient.
However, please verify that the dependency versions are current and secure, particularly the Agentuity SDK version.
#!/bin/bash # Description: Check for latest versions and security advisories for key dependencies echo "Checking Agentuity SDK version..." npm view @agentuity/sdk version echo "Checking OpenAI SDK version..." npm view openai version echo "Checking RSS parser version..." npm view rss-parser version echo "Checking for security advisories..." npm audit --audit-level=moderate --json | jq '.vulnerabilities | length'agents/Startup_News_Scraper/agentuity.yaml (1)
1-76: LGTM! Comprehensive configuration with good security practices.The Agentuity configuration is well-structured with appropriate development and deployment settings. Particularly noteworthy are the security-focused deployment flags like
--disallow-code-generation-from-strings,--no-addons, and--no-macroswhich enhance runtime security.The resource limits and file watching configuration are also appropriately set for the project needs.
frameworks/mastra/cat-fact-agent/src/agents/cat-agent/index.ts (1)
1-21: LGTM! Clean and well-structured agent implementation.This agent follows Agentuity best practices with proper type usage, comprehensive error handling, and user-friendly responses. The fallback to a default prompt and the welcome function provide good user experience. The error handling appropriately logs technical details while returning a friendly message to users.
frameworks/mastra/cat-fact-agent/package.json (1)
30-34: Verify dependency version compatibility.Ensure all dependency versions are compatible and up-to-date, particularly for the SDK and AI-related packages.
#!/bin/bash # Check for the latest versions of key dependencies echo "Checking latest versions..." npm view @agentuity/sdk version npm view @ai-sdk/openai version npm view @mastra/core version npm view openai versionframeworks/mastra/cat-fact-agent/index.ts (2)
11-25: LGTM! Comprehensive environment validation.The environment validation logic properly checks for required API keys and provides helpful error messages with runtime-specific instructions.
43-52: Good error handling in the runner.The catch block appropriately handles both Error instances and other types of exceptions, providing detailed error information before exiting.
agents/Startup_News_Scraper/.cursor/rules/sdk.mdc (1)
88-97: LGTM! Excellent best practices section.The best practices section provides valuable guidance for SDK usage, covering TypeScript, error handling, logging, and storage recommendations.
frameworks/mastra/blog-publisher/package.json (2)
29-35: Good dependency selection for blog publishing functionality.The addition of
@ai-sdk/anthropicalongside OpenAI dependencies provides flexibility for different AI models in blog publishing workflows.
6-13: Consistent script configuration across agent projects.The npm scripts follow the same pattern as other agents in the repository, ensuring consistency in development workflows.
frameworks/mastra/blog-publisher/README.md (1)
1-118: Well-structured documentation with clear setup instructions.The README provides comprehensive documentation for the Blog-Publisher agent with proper structure, clear prerequisites, and step-by-step instructions. The hierarchical multi-agent workflow is well explained.
agents/Startup_News_Scraper/index.ts (1)
43-43: Verify Node.js version requirement for import.meta.dirname.The
import.meta.dirnameis only available in Node.js 20.11+ and ESM modules. Ensure your project specifies the minimum Node.js version requirement in package.json.#!/bin/bash # Check if package.json specifies minimum Node.js version fd "package.json" --exec cat {} | jq '.engines.node // "not specified"'frameworks/mastra/cat-fact-agent/README.md (1)
1-118: Comprehensive documentation with clear agent workflow explanation.The README effectively explains the Cat-Fact-Agent's purpose, system prompt usage, and tool integration. The documentation structure is consistent with other agents in the project.
frameworks/mastra/cat-fact-agent/src/agents/cat-agent/cat.ts (2)
22-31: Well-designed system prompt with clear responsibilities.The system prompt effectively guides the agent's behavior and ensures consistent use of the cat fact tool in responses.
34-39: Agent configuration looks good with appropriate model selection.The agent is properly configured with GPT-4o-mini model and integrates the custom cat fact tool effectively.
agents/Startup_News_Scraper/src/agents/startup_scraper/fetchFeeds.ts (2)
25-34: Good error handling for RSS feed parsing.The concurrent fetching with proper error handling ensures that failures in individual feeds don't break the entire operation. The warning logs will help with debugging feed issues.
39-49: Robust filtering and data processing logic.The implementation correctly handles:
- Missing feed data with null checks
- Filtering items with required fields (title, link)
- Keyword matching in both title and content
- Limiting results per feed
- Safe non-null assertions after filtering
frameworks/mastra/cat-fact-agent/agentuity.yaml (3)
51-57: Resource allocation is appropriate for a simple agent.The memory (350Mi), CPU (500M), and disk (250Mi) allocations are reasonable for a cat fact agent. These conservative limits help prevent resource overconsumption.
36-49: Excellent security-focused deployment configuration.The deployment args include comprehensive security flags:
--no-macrosand--disallow-code-generation-from-stringsprevent code injection--no-addonsreduces attack surface--report-uncaught-exceptionimproves error visibility--no-global-search-pathslimits file system access
66-69: Verify ignore patterns align with bundler requirements.The ignore patterns include
src/**which seems counterintuitive since agents are typically in the src directory. Ensure this doesn't interfere with agent bundling.#!/bin/bash # Description: Check if similar ignore patterns exist in other agentuity.yaml files # Expected: Consistent patterns across agent projects echo "Checking ignore patterns in other agentuity.yaml files:" fd -e yaml -x grep -H -A5 -B5 "ignore:" {} \;agents/Startup_News_Scraper/n8n_workflow/workflow.json (1)
163-163: Workflow is inactive by default.The workflow is set to
"active": false, which means it won't run automatically despite having a schedule trigger.Verify if this is intentional or if the workflow should be activated for production use.
frameworks/mastra/blog-publisher/src/agents/supervisor/index.ts (2)
14-16: Good input validation and user feedback.The topic extraction and validation provide clear feedback to users when no topic is provided.
32-35: Appropriate error handling with logging.The error handling correctly logs errors for debugging while providing a user-friendly error message.
frameworks/mastra/blog-publisher/.cursor/rules/agent.mdc (1)
18-24: Example agent code follows best practices.The example demonstrates proper TypeScript usage, correct imports from the SDK, and appropriate function signature.
agents/Startup_News_Scraper/.cursor/rules/agent.mdc (3)
1-5: LGTM! Well-structured frontmatter configuration.The YAML frontmatter correctly configures the cursor rules with appropriate description, glob pattern for TypeScript agent files, and the
alwaysApply: truesetting for consistent enforcement.
9-15: Clear and comprehensive guidelines.The guidelines provide essential best practices for agent development, including proper SDK usage, naming conventions, TypeScript requirements, and logging patterns. The recommendation to use the
agentuity agent createcommand promotes consistency.
18-24: Verify the example agent implementation.The example demonstrates proper TypeScript imports and function structure. However, I notice potential issues that should be verified:
#!/bin/bash # Verify that the example agent code matches the actual SDK API ast-grep --pattern $'export default async function $FUNC_NAME($REQ: AgentRequest, $RESP: AgentResponse, $CTX: AgentContext) { $$$ }'agents/Startup_News_Scraper/src/agents/startup_scraper/index.ts (1)
7-10: LGTM! Proper SDK imports and module structure.The imports correctly use the Agentuity SDK types and follow the guidelines established in the cursor rules. The local module imports are well-organized and maintain clear separation of concerns.
frameworks/mastra/blog-publisher/index.ts (5)
3-9: LGTM! Proper global interface augmentation.The global interface declaration for adding the
isBunproperty to the NodeJS Process interface is correctly implemented and follows TypeScript best practices.
11-25: Well-implemented environment validation with clear error messaging.The validation logic properly checks for required API keys and provides helpful error messages with colored output. The conditional advice for Node.js users is a nice touch for developer experience.
27-41: Good warning system for local development.The warnings for missing
AGENTUITY_URLare helpful and provide appropriate guidance based on the runtime environment (Bun vs Node.js). The colored output enhances visibility.
43-52: Robust error handling for the runner function.The error handling properly distinguishes between Error instances and other types, provides stack traces when available, and ensures proper process exit codes.
43-43: Verify import.meta.dirname usage.The usage of
import.meta.dirnameshould be verified as it's a relatively new Node.js feature and may not be available in all environments.Is import.meta.dirname supported in Node.js and what is the minimum version required?agents/Startup_News_Scraper/README.md (2)
1-12: Excellent introduction with clear branding and purpose.The README starts with a professional header that clearly explains the project's purpose. The Agentuity branding and deploy button provide good visual appeal and immediate action items.
126-127: Verify documentation links are accessible.The documentation links should be verified to ensure they're publicly accessible and contain the expected content.
#!/bin/bash # Verify that the documentation URLs are accessible curl -I https://agentuity.dev/SDKs/javascript 2>/dev/null | head -1frameworks/mastra/blog-publisher/.cursor/rules/sdk.mdc (3)
88-96: Excellent best practices section.The best practices provide valuable guidance for developers, covering type safety, error handling, logging, and architectural considerations. This adds significant value to the documentation.
97-97: Verify documentation URL accessibility.The SDK documentation link should be verified to ensure it's publicly accessible.
#!/bin/bash # Verify that the SDK documentation URL is accessible curl -I https://agentuity.dev/SDKs/javascript/api-reference 2>/dev/null | head -1
33-37: To confirm whether both methods actually exist (or if this is just a docs issue), let’s locate the realAgentRequestdefinition and itsdatamethods:#!/bin/bash # 1. Find where AgentRequest is defined rg "export (interface|type) AgentRequest" -n # 2. If found, show its data block to check for both methods rg -A5 -B2 "AgentRequest" -n --glob "*.ts" # 3. Search globally for buffer() and binary() return signatures rg -C3 "buffer(): *Promise" -n rg -C3 "binary(): *Promise" -nagents/Startup_News_Scraper/src/agents/startup_scraper/summarize.ts (2)
4-4: Verify OpenAI API key configuration.The OpenAI client initialization relies on
process.env.OPENAI_API_KEYwithout validation. Consider adding explicit error handling if the API key is missing.-const openai = new OpenAI({ apiKey: process.env.OPENAI_API_KEY }); +const openai = new OpenAI({ + apiKey: process.env.OPENAI_API_KEY || (() => { + throw new Error('OPENAI_API_KEY environment variable is required'); + })() +});
11-15: LGTM! Clean chunk-based concurrency implementation.The array chunking logic for batching API requests is well-implemented and will help manage rate limits effectively.
frameworks/mastra/cat-fact-agent/.cursor/rules/sdk.mdc (1)
1-5: LGTM! Well-structured cursor rules metadata.The YAML frontmatter is properly configured with appropriate description, glob patterns, and conditional application settings.
| ignore: | ||
| - node_modules/** | ||
| - dist/** | ||
| - src/** |
There was a problem hiding this comment.
Remove src/** from bundler ignore.
Ignoring the entire src/ directory will also exclude your agent source files (src/agents). Update to ignore only build artifacts (e.g., node_modules and dist) but keep src/agents included.
Apply this diff:
bundler:
ignore:
- - node_modules/**
- - dist/**
- - src/**
+ - node_modules/**
+ - dist/**📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| ignore: | |
| - node_modules/** | |
| - dist/** | |
| - src/** | |
| ignore: | |
| - node_modules/** | |
| - dist/** |
🤖 Prompt for AI Agents
In frameworks/mastra/blog-publisher/agentuity.yaml around lines 66 to 69, the
bundler ignore list currently excludes the entire src/ directory, which
unintentionally excludes your agent source files in src/agents. Modify the
ignore list to remove src/** and only ignore build artifacts like
node_modules/** and dist/**, ensuring that src/agents remains included in the
bundle.
| { | ||
| "name": "cat-fact-agent", | ||
| "version": "0.0.1", | ||
| "main": "index.js", |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Inconsistent main and module entry points.
The main field points to index.js while the module field points to index.ts. This inconsistency could cause module resolution issues.
Apply this diff to align the entry points:
- "main": "index.js",
+ "main": "index.ts",Also applies to: 35-35
🤖 Prompt for AI Agents
In frameworks/mastra/cat-fact-agent/package.json at line 4 and line 35, the
"main" field points to "index.js" while the "module" field points to "index.ts",
causing inconsistency in entry points. To fix this, update the "module" field to
point to "index.js" to match the "main" field or change both fields to
consistently point to the same compiled JavaScript file, ensuring consistent
module resolution.
| const getCatFact = async () => { | ||
| const { fact } = (await fetch("https://catfact.ninja/fact").then(r => r.json())) as { fact: string }; | ||
| return fact; | ||
| }; |
There was a problem hiding this comment.
Add error handling and response validation for external API call.
The fetch operation lacks error handling and doesn't validate the API response structure, which could cause runtime failures.
const getCatFact = async () => {
- const { fact } = (await fetch("https://catfact.ninja/fact").then(r => r.json())) as { fact: string };
- return fact;
+ try {
+ const response = await fetch("https://catfact.ninja/fact");
+ if (!response.ok) {
+ throw new Error(`API request failed: ${response.status}`);
+ }
+ const data = await response.json();
+ if (!data || typeof data.fact !== 'string') {
+ throw new Error('Invalid API response format');
+ }
+ return data.fact;
+ } catch (error) {
+ console.error('Failed to fetch cat fact:', error);
+ return 'Did you know cats have been domesticated for over 4,000 years?'; // Fallback fact
+ }
};📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const getCatFact = async () => { | |
| const { fact } = (await fetch("https://catfact.ninja/fact").then(r => r.json())) as { fact: string }; | |
| return fact; | |
| }; | |
| const getCatFact = async () => { | |
| try { | |
| const response = await fetch("https://catfact.ninja/fact"); | |
| if (!response.ok) { | |
| throw new Error(`API request failed: ${response.status}`); | |
| } | |
| const data = await response.json(); | |
| if (!data || typeof data.fact !== 'string') { | |
| throw new Error('Invalid API response format'); | |
| } | |
| return data.fact; | |
| } catch (error) { | |
| console.error('Failed to fetch cat fact:', error); | |
| return 'Did you know cats have been domesticated for over 4,000 years?'; // Fallback fact | |
| } | |
| }; |
🤖 Prompt for AI Agents
In frameworks/mastra/cat-fact-agent/src/agents/cat-agent/cat.ts around lines 7
to 10, the getCatFact function lacks error handling and does not validate the
API response structure. To fix this, wrap the fetch call in a try-catch block to
handle potential network or parsing errors, and add checks to ensure the
response contains the expected 'fact' property before returning it. If the
validation fails, handle the error gracefully, possibly by returning a default
value or throwing a descriptive error.
| "guildId": { | ||
| "__rl": true, | ||
| "mode": "list", | ||
| "value": "" | ||
| }, | ||
| "channelId": { | ||
| "__rl": true, | ||
| "mode": "list", | ||
| "value": "" | ||
| }, |
There was a problem hiding this comment.
Discord configuration has empty values.
The Discord node has empty guildId and channelId values, which will prevent the workflow from working correctly.
Populate the Discord configuration with actual values:
"guildId": {
"__rl": true,
"mode": "list",
- "value": ""
+ "value": "your-discord-guild-id"
},
"channelId": {
"__rl": true,
"mode": "list",
- "value": ""
+ "value": "your-discord-channel-id"
},🤖 Prompt for AI Agents
In agents/Startup_News_Scraper/n8n_workflow/workflow.json around lines 93 to
102, the Discord node configuration has empty strings for the guildId and
channelId fields. Replace these empty values with the actual Discord guild ID
and channel ID to ensure the workflow can connect and operate correctly within
the intended Discord server and channel.
| "parameters": { | ||
| "method": "POST", | ||
| "options": {} | ||
| }, | ||
| "type": "n8n-nodes-base.httpRequest", | ||
| "typeVersion": 4.2, | ||
| "position": [ | ||
| -440, | ||
| 180 | ||
| ], | ||
| "id": "a1f3a427-c810-49d6-b3c6-dd1ce4fdda16", | ||
| "name": "HTTP Request" |
There was a problem hiding this comment.
Missing HTTP request URL configuration.
The HTTP Request node lacks a URL parameter, which means it won't know where to send the POST request to the startup scraper agent.
Add the URL parameter to the HTTP Request node:
"parameters": {
"method": "POST",
+ "url": "https://your-agentuity-deployment-url.com/agent/startup_scraper",
"options": {}
},📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| "parameters": { | |
| "method": "POST", | |
| "options": {} | |
| }, | |
| "type": "n8n-nodes-base.httpRequest", | |
| "typeVersion": 4.2, | |
| "position": [ | |
| -440, | |
| 180 | |
| ], | |
| "id": "a1f3a427-c810-49d6-b3c6-dd1ce4fdda16", | |
| "name": "HTTP Request" | |
| "parameters": { | |
| "method": "POST", | |
| "url": "https://your-agentuity-deployment-url.com/agent/startup_scraper", | |
| "options": {} | |
| }, | |
| "type": "n8n-nodes-base.httpRequest", | |
| "typeVersion": 4.2, | |
| "position": [ | |
| -440, | |
| 180 | |
| ], | |
| "id": "a1f3a427-c810-49d6-b3c6-dd1ce4fdda16", | |
| "name": "HTTP Request" |
🤖 Prompt for AI Agents
In agents/Startup_News_Scraper/n8n_workflow/workflow.json around lines 24 to 35,
the HTTP Request node is missing the URL parameter, so it does not know the
endpoint to send the POST request. Add a "url" field inside the "parameters"
object with the correct endpoint URL string to specify where the request should
be sent.
| const raw = res.choices?.[0]?.message?.content ?? '{}'; | ||
| let parsed: any = {}; | ||
| try { | ||
| parsed = JSON.parse(raw); | ||
| } catch { | ||
| parsed = {}; | ||
| } |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Improve JSON parsing error handling.
The empty catch block silently swallows parsing errors. Consider logging the parsing failure for debugging.
const raw = res.choices?.[0]?.message?.content ?? '{}';
let parsed: any = {};
try {
parsed = JSON.parse(raw);
-} catch {
+} catch (error) {
+ console.warn('Failed to parse OpenAI response as JSON:', { raw, error: error.message, url: link.url });
parsed = {};
}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const raw = res.choices?.[0]?.message?.content ?? '{}'; | |
| let parsed: any = {}; | |
| try { | |
| parsed = JSON.parse(raw); | |
| } catch { | |
| parsed = {}; | |
| } | |
| const raw = res.choices?.[0]?.message?.content ?? '{}'; | |
| let parsed: any = {}; | |
| try { | |
| parsed = JSON.parse(raw); | |
| } catch (error) { | |
| console.warn('Failed to parse OpenAI response as JSON:', { | |
| raw, | |
| error: error.message, | |
| url: link.url, | |
| }); | |
| parsed = {}; | |
| } |
🤖 Prompt for AI Agents
In agents/Startup_News_Scraper/src/agents/startup_scraper/summarize.ts around
lines 53 to 59, the JSON parsing error is caught with an empty catch block,
which hides parsing issues. Modify the catch block to log the error details
using an appropriate logger or console.error to provide visibility into parsing
failures for easier debugging.
| const prompt = `You are a venture-capital analyst writing deal briefs. | ||
|
|
||
| Read the article at: | ||
| ${link.url} | ||
|
|
||
| Return a single **JSON object** with these keys ⬇︎ | ||
| - name | ||
| - industry | ||
| - summary – 2-4 sentences, 60-120 words, plain text | ||
| - founders – null if unknown | ||
| - why_it_matters – 1-2 sentences on strategic significance, null if unknown | ||
| - funding_source – null if not mentioned`; |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Improve prompt clarity and add URL validation.
The prompt could be more explicit about expected JSON structure, and consider validating the URL before sending to OpenAI.
+ // Basic URL validation
+ if (!link.url || !link.url.startsWith('http')) {
+ throw new Error(`Invalid URL: ${link.url}`);
+ }
+
const prompt = `You are a venture-capital analyst writing deal briefs.
Read the article at:
${link.url}
-Return a single **JSON object** with these keys ⬇︎
+Return a single JSON object with exactly these keys (use null for unknown values):
+{
+ "name": "string - company name",
+ "industry": "string - industry category",
+ "summary": "string - 2-4 sentences, 60-120 words",
+ "founders": "string or null - founder names",
+ "why_it_matters": "string or null - strategic significance",
+ "funding_source": "string or null - funding details"
+}`;📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const prompt = `You are a venture-capital analyst writing deal briefs. | |
| Read the article at: | |
| ${link.url} | |
| Return a single **JSON object** with these keys ⬇︎ | |
| - name | |
| - industry | |
| - summary – 2-4 sentences, 60-120 words, plain text | |
| - founders – null if unknown | |
| - why_it_matters – 1-2 sentences on strategic significance, null if unknown | |
| - funding_source – null if not mentioned`; | |
| // Basic URL validation | |
| if (!link.url || !link.url.startsWith('http')) { | |
| throw new Error(`Invalid URL: ${link.url}`); | |
| } | |
| const prompt = `You are a venture-capital analyst writing deal briefs. | |
| Read the article at: | |
| ${link.url} | |
| Return a single JSON object with exactly these keys (use null for unknown values): | |
| { | |
| "name": "string - company name", | |
| "industry": "string - industry category", | |
| "summary": "string - 2-4 sentences, 60-120 words", | |
| "founders": "string or null - founder names", | |
| "why_it_matters": "string or null - strategic significance", | |
| "funding_source": "string or null - funding details" | |
| }`; |
🤖 Prompt for AI Agents
In agents/Startup_News_Scraper/src/agents/startup_scraper/summarize.ts around
lines 31 to 42, improve the prompt by explicitly specifying the expected JSON
format, including data types and example values for each key to enhance clarity.
Additionally, add a URL validation step before using link.url in the prompt to
ensure it is a valid URL, preventing malformed inputs from being sent to OpenAI.
| - `request.data.contentType`: Gets the content type of the request payload | ||
| - `request.data.json(): Promise<Json>`: Gets the payload as a JSON object | ||
| - `request.data.text(): Promise<string>`: Gets the payload as a string | ||
| - `request.data.buffer(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer | ||
| - `request.data.binary(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer | ||
| - `request.data.object<T>: Promise<T>`: Gets the payload as a typed object |
There was a problem hiding this comment.
Fix inconsistent method signatures.
The documentation shows inconsistent method names and signatures. Some methods use different naming patterns.
- `request.data.json(): Promise<Json>`: Gets the payload as a JSON object
- `request.data.text(): Promise<string>`: Gets the payload as a string
- `request.data.buffer(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer
-- `request.data.binary(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer
-- `request.data.object<T>: Promise<T>`: Gets the payload as a typed object
+- `request.data.binary(): Promise<ArrayBuffer>`: Gets the payload as binary data
+- `request.data.object<T>(): Promise<T>`: Gets the payload as a typed objectNote: buffer() and binary() appear to have identical return types - verify if this is correct or if they serve different purposes.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| - `request.data.contentType`: Gets the content type of the request payload | |
| - `request.data.json(): Promise<Json>`: Gets the payload as a JSON object | |
| - `request.data.text(): Promise<string>`: Gets the payload as a string | |
| - `request.data.buffer(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer | |
| - `request.data.binary(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer | |
| - `request.data.object<T>: Promise<T>`: Gets the payload as a typed object | |
| - `request.data.contentType`: Gets the content type of the request payload | |
| - `request.data.json(): Promise<Json>`: Gets the payload as a JSON object | |
| - `request.data.text(): Promise<string>`: Gets the payload as a string | |
| - `request.data.buffer(): Promise<ArrayBuffer>`: Gets the payload as a ArrayBuffer | |
| - `request.data.binary(): Promise<ArrayBuffer>`: Gets the payload as binary data | |
| - `request.data.object<T>(): Promise<T>`: Gets the payload as a typed object |
🤖 Prompt for AI Agents
In frameworks/mastra/cat-fact-agent/.cursor/rules/sdk.mdc between lines 32 and
37, the method signatures for request.data are inconsistent, with some using
parentheses and others not, and buffer() and binary() having identical return
types. Standardize all method signatures to use parentheses for methods
returning promises, confirm if buffer() and binary() are distinct or redundant,
and update the documentation to reflect accurate method names and return types
consistently.
| - `context.kv.get(name, key)`: Retrieves a value | ||
| - `context.kv.set(name, key, value, params)`: Stores a value with optional params (KeyValueStorageSetParams) | ||
| - `context.kv.delete(name, key)`: Deletes a value |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Add parameter type information for KV storage methods.
The key-value storage methods lack parameter type information, particularly for the params parameter.
-- `context.kv.set(name, key, value, params)`: Stores a value with optional params (KeyValueStorageSetParams)
+- `context.kv.get(name: string, key: string): Promise<any>`: Retrieves a value
+- `context.kv.set(name: string, key: string, value: any, params?: KeyValueStorageSetParams): Promise<void>`: Stores a value with optional params
+- `context.kv.delete(name: string, key: string): Promise<void>`: Deletes a valueCommittable suggestion skipped: line range outside the PR's diff.
🤖 Prompt for AI Agents
In frameworks/mastra/cat-fact-agent/.cursor/rules/sdk.mdc around lines 66 to 68,
the documentation for context.kv methods is missing parameter type details,
especially for the params argument in context.kv.set. Update the method
descriptions to include explicit parameter types, such as specifying that params
is of type KeyValueStorageSetParams, to improve clarity and usability of the SDK
documentation.
| - `context.vector.upsert(name, ...documents)`: Inserts or updates vectors | ||
| - `context.vector.search(name, params)`: Searches for vectors | ||
| - `context.vector.delete(name, ...ids)`: Deletes vectors |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Add parameter type information for vector storage methods.
Similar to KV storage, vector storage methods need parameter type documentation.
-- `context.vector.upsert(name, ...documents)`: Inserts or updates vectors
-- `context.vector.search(name, params)`: Searches for vectors
-- `context.vector.delete(name, ...ids)`: Deletes vectors
+- `context.vector.upsert(name: string, ...documents: VectorDocument[]): Promise<void>`: Inserts or updates vectors
+- `context.vector.search(name: string, params: VectorSearchParams): Promise<VectorSearchResult[]>`: Searches for vectors
+- `context.vector.delete(name: string, ...ids: string[]): Promise<void>`: Deletes vectors📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| - `context.vector.upsert(name, ...documents)`: Inserts or updates vectors | |
| - `context.vector.search(name, params)`: Searches for vectors | |
| - `context.vector.delete(name, ...ids)`: Deletes vectors | |
| - `context.vector.upsert(name: string, ...documents: VectorDocument[]): Promise<void>`: Inserts or updates vectors | |
| - `context.vector.search(name: string, params: VectorSearchParams): Promise<VectorSearchResult[]>`: Searches for vectors | |
| - `context.vector.delete(name: string, ...ids: string[]): Promise<void>`: Deletes vectors |
🤖 Prompt for AI Agents
In frameworks/mastra/cat-fact-agent/.cursor/rules/sdk.mdc around lines 74 to 76,
the vector storage methods lack parameter type documentation. Add detailed
parameter type information for the methods context.vector.upsert,
context.vector.search, and context.vector.delete, similar to how it is done for
KV storage methods, specifying the expected types for name, documents, params,
and ids to improve clarity and usability.
2 participants
Summary by CodeRabbit
New Features
Documentation
Configuration
Workflow Automation