training-plan

Manage concept files generated during TTT sessions -- list, search, path, link

Create a customized training plan from a learning goal using Test-Teach-Test methodology

Run a Test-Teach-Test learning session for a sub-goal from your training plan

Create a Socratic training plan where each concept arises from a discovery question and the conclusion feels inevitable

Show cross-plan learning analytics -- completion rates, strengths, gap patterns, and session history

Run autonomous QA/UX test of the training-plan plugin -- creates plan, simulates learner sessions, produces compliance and effectiveness report

Use this agent to clarify terminology during TTT test phases without leaking answers. The agent classifies terms as context vocabulary, tested knowledge, or missing prerequisites, then provides appropriate responses. <example> Context: Learner encounters an unfamiliar term during an Initial Test user: "What does 'greenfield' mean in this question?" assistant: "I'll use the term-clarifier agent to provide a definition without revealing test answers." <commentary> The term-clarifier receives only the term, the CTQ list, and the principle list — never the test answer. It classifies the term and provides a safe clarification. </commentary> </example>

Use this agent when the user needs to create a training plan, run a learning session, or work with the Test-Teach-Test methodology for any learning goal. <example> Context: User wants to learn a new topic user: "I want to learn cost estimation for EPC projects" assistant: "I'll use the training-planner agent to decompose this goal and build your training plan." <commentary> New training plan creation requires goal decomposition and session blueprint design -- the training-planner orchestrates this full pipeline. </commentary> </example> <example> Context: User wants a Socratic training plan user: "Create a Socratic training plan for deriving quantum mechanics from spectra" assistant: "I'll use the training-planner agent to build a Socratic training plan with discovery questions and guidance chains." <commentary> Socratic training plan uses the same goal decomposition as /create-training-plan plus a Socratic Discovery layer (Discovery Questions, Socratic Guidance Chains, designed wrong paths, reconstruct-the-reasoning tests). </commentary> </example> <example> Context: User has a training plan and wants to practice user: "Run session SG-3 on unit-rate estimation" assistant: "I'll use the training-planner agent to run a TTT session for that sub-goal." <commentary> Running a TTT session requires reading the plan, presenting practical cases, evaluating responses, and adapting -- the training-planner handles the full interactive loop. </commentary> </example> <example> Context: User wants to see their learning progress user: "How am I doing across all my training plans?" assistant: "I'll use the training-planner agent to show your training status." <commentary> Cross-plan analytics require reading the learner profile and all training plans -- the training-planner handles this via /training-status. </commentary> </example> <example> Context: User wants to browse concept files from sessions user: "Show me the concepts I've learned so far" assistant: "I'll use the training-planner agent to list your concept library." <commentary> The concept library tracks concept files generated during Teach phases -- the training-planner handles this via /concept-library. </commentary> </example>

Use this agent to run autonomous QA/UX testing of the training-plan plugin. Give it a learning goal and it will create a training plan, simulate a learner through all sessions with varied performances, then inspect every artifact and produce a comprehensive compliance and effectiveness report. <example> Context: Developer wants to test the full plugin pipeline user: "Run user testing on thermodynamic entropy" assistant: "I'll use the user-testing agent to create a plan, simulate sessions with varied performances, and produce a QA report." <commentary> Full autonomous test: the agent creates the plan, simulates a learner through all sub-goals (zero-gap pass, partial failure, full failure with loop, "not familiar", clarification request), then inspects all artifacts against specs and writes a QA report. </commentary> </example> <example> Context: Developer wants to QA-inspect existing session files user: "Inspect my existing training plan and sessions for spec compliance" assistant: "I'll use the user-testing agent in inspect mode to check your existing files against all plugin specifications." <commentary> Inspect-only mode: skips plan creation and simulation, reads existing plan files, transcripts, learner profile, and concept files, then produces the QA report. </commentary> </example> <example> Context: Developer wants to test with a specific performance profile user: "Run user testing on SQL optimization -- make all sessions fail" assistant: "I'll use the user-testing agent with a custom performance profile where every session results in failure." <commentary> Custom performance profile: the developer can override the default varied-performance profile to stress-test specific code paths (e.g., all failures, all passes, all "not familiar"). </commentary> </example>

Concept explanation methodology for TTT sessions. Use when generating teaching content during the Teach phase, when a learner requests a deep-dive into a taught principle, when gaps involve distinguishing two concepts (comparison), or when generating concept maps and CTQ mastery criteria.

Concept map generation for training plan topics. Use when the user asks to map concepts, visualize concept relationships, see how topics connect, generate a concept map for a training plan argument, create a knowledge map, show dependencies between concepts, or map all concepts in a domain with links to the training plan's main concepts.

Gap-targeted teaching content generation for training sessions. Use when generating teaching material during the Teach phase of a TTT session, when the user asks to explain a concept, teach me about a topic, what did I get wrong, fill a learning gap, when a gap involves distinguishing two concepts, or when the learner requests a deep-dive into a taught principle.

Learning goal decomposition into SMART sub-goals. Use when the user asks to create a training plan, break down a learning goal, decompose objectives, define sub-goals, structure a learning path from a high-level goal, or identify what sub-topics need to be learned for a given goal.

Learner profile management and cross-plan analytics. Use when updating the learner profile after a session, reading the profile to inform plan creation, generating training status reports, analyzing gap patterns, identifying learner strengths, or computing completion statistics across multiple training plans.

Test-Teach-Test session design and execution. Use when the user asks to run a session, start a learning session, test me on a topic, practice a skill, assess my knowledge, design a TTT learning cycle, or execute an adaptive test-teach-test loop for a sub-goal.