Mcporter

My agent skills.

Install

npx -y skills add zlliang/skills --skill mcporter

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Inspect and call MCP server tools from the command line with MCPorter. Use when a task needs an MCP tool (search, docs, integrations) or when checking which MCP servers and tools are available.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

3.9 KB, 929 tokens by cl100k_base, as published. Nobody here has run it

MCPorter (`mcporter`)

MCPorter is a CLI for working with MCP servers. Use it to discover configured servers, read their tool signatures, and call tools directly from the shell.

The two commands that matter most are list (discover and inspect) and call (invoke). Always list before you call so arguments match the real schema.

For exact flags, run mcporter <command> --help. MCPorter discovers servers from its config — the global config at ~/.mcporter/mcporter.json, a project config at ./config/mcporter.json, or an explicit --config <path> — plus editor imports (Cursor, Claude Code, Codex, etc.). Use mcporter config list or mcporter list --verbose to see where each server comes from.

Workflow

# 1. See configured servers and their health
mcporter list

# 2. Read one server's tools as TypeScript-like signatures
mcporter list <server>

# 3. Call a tool using the signature from step 2
mcporter call '<server>.<tool>(arg: "value", count: 5)'

`list` — discover and inspect

mcporter list                             # All servers, with live status
mcporter list <server>                    # Tools for one server, as signatures
mcporter list <server> --schema           # Raw JSON schema for each tool
mcporter list <server> --json             # Machine-readable, full tool metadata
mcporter list <server> --all-parameters   # Reveal hidden optional params

Single-server output reads like a TypeScript header: a doc comment per tool, a function name(...) signature with required params first and optional ones marked ?, plus an Examples: block written in the function-call syntax below. Optional params beyond the first few are hidden unless you pass --all-parameters or --schema.

function web_search_exa(query: string, numResults?: number);

`call` — invoke a tool

Recommended: function-call syntax. It mirrors the signature from mcporter list, is self-documenting, handles nested objects/arrays, and avoids shell-flag ambiguity. Wrap the whole argument in single quotes.

mcporter call 'exa.web_search_exa(query: "best React state libraries", numResults: 5)'
mcporter call 'linear.create_comment(issueId: "ENG-123", body: "Looks good!")'

Named arguments are preferred; unlabeled values map to schema order.
Use --output json for structured results; connection/auth/offline/http failures emit a { server, tool, issue } envelope for scripting.

mcporter call 'exa.web_search_exa(query: "...")' --output json

Other argument styles (when the recommended one is awkward)

# Shell-friendly key/value pairs (good in scripts)
mcporter call linear.list_issues team=ENG limit:5

# Prebuilt JSON payload (or `--json -` to read from stdin)
mcporter call linear.create_issue --json '{"title":"Bug","team":"ENG"}'

# Read a long string argument from a file
mcporter call linear.create_comment issueId=LNR-123 [email protected]

Ad-hoc servers (not in config)

Point list/call at an endpoint directly, no config edit needed:

mcporter list https://mcp.linear.app/mcp
mcporter call 'https://www.shadcn.io/api/mcp.getComponent(component: "vortex")'
mcporter list --stdio "bun run ./server.ts" --env TOKEN=xyz

Bare domains assume https://; plain http:// needs --allow-http. Persist a tried endpoint with --persist <config> to make its short name reusable.

Notes

Pass secrets via environment variables, e.g. LINEAR_API_KEY=… mcporter call 'linear.search_documentation(query: "automations")'.
Other commands exist (auth, config, resource, daemon, generate-cli, emit-ts); run mcporter --help when you need them.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most mcp tooling skills give in 929 tokens

Counted across 638 of the 750 authors here whose files we hold, read 2026-08-07

Create ten complex or independent read-only evaluation questionsin 69 of 638, across 15 files
Test servers using MCP Inspectorin 61 of 638, across 19 files
Provide actionable error messages with specific next stepsin 54 of 638, across 12 files
Prioritize comprehensive API coverage over specific workflows or workflow toolsin 54 of 638, across 12 files
Use TypeScript and Streamable HTTP for remote servers or clientsin 54 of 638, across 8 files
Define structured output schemas where possiblein 50 of 638, across 8 files
Use Zod or Pydantic for input schemasin 47 of 638, across 5 files
Fetch MCP specification pages with markdown suffixin 46 of 638, across 4 files
Load framework documentation using WebFetchin 45 of 638, across 3 files
Verify each evaluation answer independentlyin 45 of 638, across 3 files
Implement API client with authentication and paginationin 45 of 638, across 3 files
Define input schemas with validationin 27 of 638, across 9 files

Said here and by no other author read

always list before calling a tool
wrap function-call arguments in single quotes
use named arguments

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.