Mastery Pathing

Knowledge Graph, Mastery & Next-Topic Planner

mastery_pathing

Read-only

Adaptive-learning engine. From a curriculum's prerequisite map (topics + prerequisites) and learners' answers (learner, topic, question, correct, time) it returns: graph checks (loops with their path, undefined prerequisites, redundant links, levels, a valid teaching order); each learner's mastery of each topic (Bayesian Knowledge Tracing fitted per topic, standard settings at cold start; Elo rating for chance of a right answer and question difficulty); the knowledge frontier (topics ready to learn: not mastered, all prerequisites mastered, Knowledge Space Theory) with the next best topic and why; review-first when stuck; gaps blocking a goal and the path to it with time; shaky foundations; curriculum checks from data (topics that behave like two skills, practice with no learning gain, prerequisite links the data contradicts or suggests). Learners are never ranked. Decision cards first. Up to 5,000 answers; larger: mastery_pathing_get_engine.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "mastery_pathing"
  }
}

Input Schema

Name	Required	Description	Default
`goal`	No	Target topic for gap diagnosis and the path to it.
`method`	No	auto/bkt = knowledge tracing (fitted or standard); rating = Elo-only mastery.	auto
`topics`	No	Topics: topic_id, prerequisites ('A; B' or list), minutes, title.
`learners`	No	Plan for these learner codes only (others still inform the topic settings).
`goal_only`	No	Only recommend topics on the way to the goal.
`responses`	No	One row per answer: learner_id (code), topic, question_id, correct (0/1), timestamp.
`model_check`	No
`prerequisites`	No	Optional links instead of a prerequisites column: {from, to} or [from, to].
`slip_guess_cap`	No
`stuck_attempts`	No
`max_split_tests`	No
`max_ready_listed`	No
`rating_mastery_p`	No
`split_bic_margin`	No
`topic_parameters`	No	Stored topic settings from an earlier run: topic, start_known, learn_rate, slip, guess.
`mastery_threshold`	No
`min_answers_topic`	No
`min_learners_edge`	No
`min_learners_topic`	No
`edge_violation_rate`	No
`infer_prerequisites`	No
`max_learners_listed`	No
`max_suggested_links`	No
`rating_min_attempts`	No
`max_questions_listed`	No

Output Schema

Name	Required	Description
`graph`	No
`alerts`	No
`status`	Yes	1 = ok, 0 = nothing could be planned
`topics`	No
`message`	No
`learners`	No
`metadata`	Yes
`decisions`	Yes	Plain-language cards, priority order. Relay these first.
`assumptions`	No
`graph_check`	No
`model_check`	No
`data_quality_report`	No
`question_difficulty`	No

Schema Changelog

Changes observed during successful MCP inspections.

First observedOct 23, 2026

TDQS

B3.2/5.0

Behavior3/5

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

Annotations already declare readOnlyHint=true, destructiveHint=false, openWorldHint=false, so the safety profile is covered. The description does add real behavioral context beyond them: no learner ranking, cold-start use of standard BKT settings, decision cards emitted first, and explicit size limits. It does not state what happens to supplied topic_parameters versus fitted ones, whether results are cached/persisted, or runtime/cost expectations for a 5,000-row fit.

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

Conciseness3/5

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

It is front-loaded and information-dense with no filler sentences, but it is delivered as one sprawling semicolon-and-parenthetical paragraph covering returns that the output schema already documents. The sibling reference and the size limit, which are the operationally important bits, sit buried at the end rather than up front.

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

Completeness2/5

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

For a 25-parameter tool with 0 required parameters and 32% schema coverage, the agent needs at minimum a signal on which knobs matter and what a minimal viable call looks like; the description gives none. It also spends most of its length re-explaining return values that the output schema already covers. It is not misleading, but a caller would have to guess at every tuning parameter.

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

Parameters2/5

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

Schema description coverage is only 32% across 25 parameters, so the description must compensate and largely does not: model_check, slip_guess_cap, stuck_attempts, mastery_threshold, split_bic_margin, edge_violation_rate, min_* thresholds, and the max_*_listed caps are all silent. It does gloss the input shapes ('topics + prerequisites', answers as learner/topic/question/correct/time) and the stored topic_parameters, but that only covers a handful of the 25. An agent has no way to know that min_answers_topic=50 gates whether a topic is fitted at all.

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

Purpose4/5

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

The description names a specific domain (adaptive-learning engine over a curriculum prerequisite map plus answer logs) and enumerates the exact output families: graph checks, per-learner BKT/Elo mastery, knowledge frontier with next best topic, gap paths, and data-driven curriculum checks. It also distinguishes itself from the sibling by reserving mastery_pathing_get_engine for >5,000 answers. It falls short of 5 only because the opening 'Adaptive-learning engine' is abstract and the real purpose has to be inferred from the return list.

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?

The 5,000-answer boundary explicitly routes the agent to mastery_pathing_get_engine, which is a concrete when-to-use-this-vs-alternative rule. It also signals output conventions ('Decision cards first', 'review-first when stuck', 'Learners are never ranked') that shape how results should be consumed. There is no guidance on when this tool is inappropriate beyond the size limit, or on data prerequisites (e.g., minimum answers per topic).

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

All Mastery Pathing tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.