Skip to main content
Glama

analyze_game_log

Replay a pasted duels.ink game log turn by turn to see why you won or lost, and get concrete tips for your next match based on your deck list.

Instructions

Review a finished (or conceded) duels.ink game from its pasted game log: what happened turn by turn, why the game was won or lost, and concrete tips for next time.

How the player gets the log: in duels.ink, open the in-game log panel and copy all of it. Paste it as log exactly as copied. It works with one event per line or with everything run together on one line (some copies lose the line breaks). "You" in the log is the player asking.

What it does: replays the log to track lore, ink, your hand (you see your own draws), both boards and every banish with its cause. It undoes any "free undo" steps first. Reports:

  • result, turns, on the play/draw, mulligans;

  • a per-turn table of your ink, plays (ink paid vs. available), songs, quests and the lore race;

  • the opponent's inks, cards seen, and the closest known tournament list (from get_meta's bundled decklists, same inks only);

  • key moments (multi-kill turns, top opposing questers);

  • tips: damage-sweeper exposure, board collapses to challenges, questing into a challenge, threats left alive too long, missed ink, unspent ink, ink choice, losing challenges, slow starts.

Deck-specific tips need the player's deck. Pass it as deck_list ("4x Card Name" per line, the duels.ink import format) or deck_path (a file containing such a list, e.g. a deck .md with a duels.ink import block; the last such block is used). If neither is given and you don't already know which deck the player was on, ask them which deck they played before (or after) calling this, since the deck comparison flags inked low-count payoffs, key 3–4-ofs never seen, and log cards that aren't in the list.

Limits: the opponent's hand is hidden. Activated-ability costs and ink- drop payments aren't itemised in duels.ink logs, so ink math counts card costs paid. Tips are heuristics. Present them as things to check, not verdicts.

Args: log: The full duels.ink game log text. deck_list: Optional deck list the player used ("4x Card Name" per line). deck_path: Optional path to a file containing the deck list.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "log": "<log>"
    },
    "name": "analyze_game_log"
  }
}

Input Schema

NameRequiredDescriptionDefault
logYes
deck_listNo
deck_pathNo

Output Schema

NameRequiredDescriptionDefault
resultYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. Addedv2.7.0

TDQS

A4.3/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full behavioral burden, and it does disclose real limits: hidden opponent hand, activated-ability and ink-drop costs not itemized, heuristic tips that should be presented as checks rather than verdicts. It also explains internal behavior (free undos stripped first, hand tracking, banish causes). It does not discuss cost, latency, or failure modes for malformed logs, so it is strong but not exhaustive.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with purpose, then how to get the log, what it does, deck requirements, and limits — a sensible order with headers doing real work. The bulleted report list is long and partly overlaps what an output schema would convey, but each item is specific rather than filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a three-parameter analysis tool with no annotations, the definition covers input acquisition, optional-parameter handling, output contents, and known blind spots. An agent has everything needed to call it correctly and to set expectations about what the analysis can and cannot claim.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, and the description fully compensates: `log` must be pasted exactly as copied and tolerates lost line breaks, `deck_list` uses the duels.ink import format ("4x Card Name" per line), and `deck_path` points to a file with such a list where the last import block is used. It also explains why the deck matters (inked low-count payoffs, 3–4-ofs never seen, cards not in the list).

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource (review a finished duels.ink game from its pasted log) plus the three deliverables: turn-by-turn account, cause of win/loss, and next-time tips. The scope is narrow enough that it is clearly distinct from analyze_deck or get_meta, which it references as dependencies rather than duplicating.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives clear use conditions — a finished or conceded game, with the log copied from the in-game log panel — and a concrete rule for the deck parameters, including instructing the agent to ask the player which deck was played if it is unknown. It does not name a sibling alternative for, say, analyzing a deck on its own, so it stops short of full when-not/alternative routing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.