MCP-native knowledge graph with code/cloud/CI indexing, structural AST search, persistent reasoning (think/charge/recall), and cross-graph traversal. Local-first, optionally cloud-synced via Fulminate.
Agent Claude Code
Knowledge graph-powered causal explorer. Authors thought clusters that explain WHY systems exist and behave the way they do, weaving evidence across code, cloud, practice, and knowledge graphs. Distinct from researcher (which describes WHAT exists and HOW it works).
Agent Claude Code
Knowledge graph-powered plan implementer. Follows plan steps sequentially, updates status in the graph, verifies success criteria, and records findings. Use after a plan has been created and approved.
Agent Claude Code
Knowledge graph-powered adversarial infrastructure-change reviewer. Audits an infrastructure changeset BEFORE any infra command runs (deploy / apply / upgrade / provision / image roll), with the skepticism of a senior engineer who has been burned by config that looked fine and broke on contact with reality.…
Agent Claude Code
Hand-crafted pattern synthesis for the practice/design-patterns.bin library graph. Reads an authoritative source (book / public catalog / reference site) and produces high-fidelity granular nodes (pattern + usecase + example + reference). You are the hydrate stage of the /ingest-patterns pipeline — the stages before…
Agent Claude Code
Knowledge graph-powered adversarial plan reviewer. Audits plans before implementation with the skepticism of a senior engineer reviewing a subordinate's work. Walks every step, verifies every claim, classifies every proposed unit, and surfaces flaws across reuse, architecture, performance, can-kicking…
Agent Claude Code
Knowledge graph-powered implementation planner. Researches the codebase and existing decisions first, then creates structured phased plans with success criteria. Use when starting a new feature, refactor, or multi-step task.
Agent Claude Code
Knowledge graph-powered researcher. Uses semantic search, code graph traversal, and knowledge nodes (decisions, findings, plans) to deeply investigate topics. Faster and more thorough than grep/glob.
Agent Claude Code
Knowledge graph-powered test plan designer. Researches what needs testing, discusses scope and criteria interactively, then creates structured test plans. Use when defining what to test and how.
Agent Claude Code
Knowledge graph-powered test executor. Runs test plans step by step, executes test commands, and records pass/fail/skip results. Read-only — reports results without fixing failures.