kata

name	kata
version	1.0.0
description	Agentic engineering katas: proven prompt/loop/tooling patterns from JM Labs. Topics: adaptive-investigation, builtin-tool-selection, confidence-stratified-sampling, context-dilution-mitigation, critical-self-correction, custom-commands-skills, defensive-structured-extraction, deterministic-agent-loop, false-positive-criteria, fewshot-edge-calibration, headless-code-review, hierarchical-claude-memory, hub-and-spoke-isolation, human-handoff-protocol, independent-reviewer-multipass, mcp-server-configuration, mcp-structured-errors, message-batch-processing, multiagent-error-propagation, multipass-prompt-chaining, path-conditional-rules, persistent-scratchpad, plan-mode-exploration, posttooluse-normalization, prefix-caching, pretooluse-guardrails, provenance-preservation, session-resume-fork, tool-description-quality, validation-retry-feedback.
params	{"topic":{"enum":["adaptive-investigation","builtin-tool-selection","confidence-stratified-sampling","context-dilution-mitigation","critical-self-correction","custom-commands-skills","defensive-structured-extraction","deterministic-agent-loop","false-positive-criteria","fewshot-edge-calibration","headless-code-review","hierarchical-claude-memory","hub-and-spoke-isolation","human-handoff-protocol","independent-reviewer-multipass","mcp-server-configuration","mcp-structured-errors","message-batch-processing","multiagent-error-propagation","multipass-prompt-chaining","path-conditional-rules","persistent-scratchpad","plan-mode-exploration","posttooluse-normalization","prefix-caching","pretooluse-guardrails","provenance-preservation","session-resume-fork","tool-description-quality","validation-retry-feedback"],"required":true,"infer":"from user request; ask only if ambiguous"},"depth":{"enum":["quick","deep"],"default":"quick"}}
routes	{"adaptive-investigation":"references/adaptive-investigation.md","builtin-tool-selection":"references/builtin-tool-selection.md","confidence-stratified-sampling":"references/confidence-stratified-sampling.md","context-dilution-mitigation":"references/context-dilution-mitigation.md","critical-self-correction":"references/critical-self-correction.md","custom-commands-skills":"references/custom-commands-skills.md","defensive-structured-extraction":"references/defensive-structured-extraction.md","deterministic-agent-loop":"references/deterministic-agent-loop.md","false-positive-criteria":"references/false-positive-criteria.md","fewshot-edge-calibration":"references/fewshot-edge-calibration.md","headless-code-review":"references/headless-code-review.md","hierarchical-claude-memory":"references/hierarchical-claude-memory.md","hub-and-spoke-isolation":"references/hub-and-spoke-isolation.md","human-handoff-protocol":"references/human-handoff-protocol.md","independent-reviewer-multipass":"references/independent-reviewer-multipass.md","mcp-server-configuration":"references/mcp-server-configuration.md","mcp-structured-errors":"references/mcp-structured-errors.md","message-batch-processing":"references/message-batch-processing.md","multiagent-error-propagation":"references/multiagent-error-propagation.md","multipass-prompt-chaining":"references/multipass-prompt-chaining.md","path-conditional-rules":"references/path-conditional-rules.md","persistent-scratchpad":"references/persistent-scratchpad.md","plan-mode-exploration":"references/plan-mode-exploration.md","posttooluse-normalization":"references/posttooluse-normalization.md","prefix-caching":"references/prefix-caching.md","pretooluse-guardrails":"references/pretooluse-guardrails.md","provenance-preservation":"references/provenance-preservation.md","session-resume-fork":"references/session-resume-fork.md","tool-description-quality":"references/tool-description-quality.md","validation-retry-feedback":"references/validation-retry-feedback.md"}

kata

Router skill: one entry, 30 agentic-engineering playbooks. Resolve topic, Read EXACTLY ONE playbook from routes:, apply it. [DOC]

When to use

Trigger when the request maps to a known agentic pattern — building a Claude agent loop, MCP server, hook, structured extraction, code-review harness, memory hierarchy, multi-agent topology, or prompt/sampling/context tactic — and you want the proven JM Labs recipe instead of improvising. [INFERENCIA] Do NOT use as a generic chat or to answer questions that no playbook covers; if no topic fits, say so and route the user elsewhere rather than guessing. [DOC]

Inputs

• topic (required): one of the 30 routes: keys. Infer from the request; ask only when two topics are genuinely plausible. [DOC]
• depth (default quick): quick → essentials + the validation gate only; deep → apply exhaustively, verifying at each step. [DOC]

Procedure

1. Map request → topic. If ambiguous, present the 2 closest keys and ask. [DOC]
2. Read EXACTLY ONE playbook (its routes: path). Never load the whole cluster — that dilutes context and defeats hub-and-spoke isolation. [INFERENCIA]
3. Execute along the spine: Discover → Analyze → Execute → Validate. [DOC]
4. Apply the playbook's own acceptance criteria before declaring done. [DOC]

Validation gate (acceptance)

• Exactly one playbook was read; topic matches the user's actual intent. [DOC]
• Output follows the chosen playbook's structure, not improvised prose. [DOC]
• Every non-obvious claim carries an evidence tag from the kit (Alfa/bracket) family per ../../references/verification-tags.md — never mix tag families. [DOC]
• Constitution v6.0.0 gates honored: enforcement, evidence tags, script-first (prefer a script over manual steps when one exists). [DOC]
• Score the result with assets/quality-rubric.json and run assets/routing-checklist.md before declaring done (see assets/README.md). [DOC]

Anti-patterns

• Loading several playbooks "to compare" — pick one; re-route if wrong. [DOC]
• Guessing a topic silently when the request is ambiguous. [DOC]
• Answering from memory of a pattern instead of reading its current playbook — recipes drift; the file is the source of truth. [INFERENCIA]
• Emitting quick depth but skipping the validation gate. [DOC]

Self-correction

If mid-task the evidence contradicts the chosen topic (wrong failure mode, the playbook's preconditions don't hold), STOP, name the mismatch, and re-resolve topic — do not force-fit the original playbook. [INFERENCIA]