Find agent skills by outcome

const truncatedData = data.slice(0

Math.max(1

Data

- Return complete

structured data suitable for programmatic processing

const data = await apiRequest(params.limit

params.offset);

content: [{ type: text

text: textContent }]

textContent = JSON.stringify(output

null

This tool searches across all user profiles in the Example platform

supporting partial matches and various search filters. It does NOT create or modify users

│ └── constants.ts # Shared constants (API_URL

CHARACTER_LIMIT

Productivity

- The `register*` methods provide better type safety

automatic schema handling

- DO NOT use: Old deprecated APIs such as `server.tool()`

`server.setRequestHandler(ListToolsRequestSchema

content: [{ type: text

text: JSON.stringify(output) }]

Productivity

2. Load each task instruction and in parallel using the MCP server and tools

identify the correct answer by attempting to solve the task YOURSELF

Productivity

Remember to parallelize solving tasks to avoid running out of context

then accumulate all answers and make changes to the file at the end.

After inspecting the content

create 10 human-readable questions:

<question>Look for pull requests that modified files in the /api directory and were merged between January 1 and January 31

2024. How many different contributors worked on these PRs?</question>

- Goal: identify specific content (e.g.

users

After understanding the API and tools

USE the MCP server tools:

- Questions must be designed so there is a single

clear answer

- Look at old content (e.g.

conversations that have ended

return { content: [{ type: text

text: result }] };

- Store API keys in environment variables

never in code

Data

- Return pagination metadata: Include `has_more`

`next_offset`/`next_cursor`

3. Be action-oriented: Start with verbs (get

list

- Python: `{service}_mcp` (e.g.

`slack_mcp`)

- MCP Protocol: Start with sitemap at `https://modelcontextprotocol.io/sitemap.xml`

then fetch specific pages with `.md` suffix