Why MCP won — the protocol wars of 2024-25 (step 8/8) · mcp

Checkpoint

One last thing before we move on. Same surface as a write step — but the lesson doesn't complete until this passes.

Final drill. Synthesize the postmortem method into one function: rank_protocols(protocols) that takes a list of protocol specs and returns a DICT mapping each protocol's name to its verdict, sorted by score DESCENDING (highest scoring first).

Each protocol is a dict with:

name: string
open_spec: bool
multi_client: bool
low_publisher_bar: bool
primitive_not_marketplace: bool

Score each protocol the same way as step 07 (25 points per True signal). Verdict thresholds:

score >= 75: "will-win"
score >= 50: "maybe"
score >= 25: "niche"
score < 25: "dead-on-arrival"

On a tie, preserve the order the protocols appeared in the input (Python's sorted is stable).

Five protocols run. Expected output:

MCP: will-win
LangChain Tools: maybe
GitHub Copilot Skills: maybe
OpenAI Plugins: dead-on-arrival
ChatGPT Actions: dead-on-arrival