By CAPHTECH
CaseGraph CLI integration for Claude Code. Skills for workspace reading/authoring/analysis, cg-driven workflow orchestration, GraphPatch proposals, and external integrations.
Use when integrating CaseGraph with external tools (markdown importer, optional sinks like Todoist / GitHub Issues, workers such as shell / local-llm / code-agent) or when verifying or recovering storage. Trigger on references to `cg import markdown`, `cg sync push|pull`, `cg worker run`, `cg validate`, `cg validate storage`, `cg cache rebuild`, `cg events verify|export`, `cg migrate`, `env_allowlist`, `approval_policy`, or phrases like "import markdown", "sync pull", "run a worker", "rebuild the cache", "verify events", "run migration". For day-to-day graph reading or editing use casegraph; for AI-authored graph changes use casegraph-patch.
Use when proposing AI-driven graph changes to a CaseGraph workspace as a `GraphPatch`. Trigger on references to `cg patch validate|review|apply`, `GraphPatch`, `base_revision`, `patch.applied`, generator metadata, the `casegraph-patch` fenced-block convention, or phrases like "propose as a patch", "wrap this change into a patch", "base_revision is stale", "apply the patch". Encodes ADR-0003 "AI does not own state". For direct manual edits use casegraph; for importer/sink/worker/recovery use casegraph-integrate.
Use when reading, manually authoring, or analyzing a CaseGraph workspace through the `cg` CLI. Trigger on references to `.casegraph/`, `cg case`, `cg node`, `cg edge`, `cg task`, `cg frontier`, `cg blockers`, `cg analyze`, `cg case view`, or phrases like "show the case", "what's blocked", "analyze impact", "create a case", "show the frontier". Do NOT use for AI-proposed graph changes (use casegraph-patch) or for importer/sink/worker/storage recovery (use casegraph-integrate).
Use when the user wants to review AI-performed work on a CaseGraph case. Trigger on phrases like "case <id> をレビューして", "AI がやった作業を確認して", "review the case", "audit this case", "証跡をチェックして", or whenever a structured read-only inspection of task completion, evidence integrity, decision traceability, and event-log trustworthiness is needed.
Use when the user wants multi-step work managed through CaseGraph instead of ad hoc chat state. Trigger on phrases like "manage this in cg", "drive this from the case", "record evidence for compaction", "resume from cg", "verify before close", or whenever durable checkpoints, verification, and guarded closure are needed for implementation, docs, investigation, or review work.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
CaseGraph is a local-first, CLI-first case graph for work that has dependencies, waits, evidence, and AI-assisted patch proposals.
Instead of treating work as a flat todo list, CaseGraph models it as a graph with:
depends_on, waits_for, and contributes_to edgesGraphPatchThe source of truth stays local. AI does not own state; it proposes patches.
0.1.1Apache-2.0Current implementation includes:
shell, code-agent, local-llm)impact, critical-path, slack, and bottlenecksPrerequisites:
npm install -g @caphtech/casegraph-cli
export WORKSPACE="$(mktemp -d /tmp/casegraph-demo.XXXXXX)"
cg --workspace "$WORKSPACE" init --title "CaseGraph Demo"
cg --workspace "$WORKSPACE" case new --id release-demo --title "Release demo"
cg --workspace "$WORKSPACE" node add --case release-demo --id goal_release_demo --kind goal --title "Release demo ready"
cg --workspace "$WORKSPACE" node add --case release-demo --id task_write_notes --kind task --title "Write release notes" --state todo
cg --workspace "$WORKSPACE" node add --case release-demo --id task_publish --kind task --title "Publish build" --state todo
cg --workspace "$WORKSPACE" edge add --case release-demo --id edge_publish_depends_notes --type depends_on --from task_publish --to task_write_notes
cg --workspace "$WORKSPACE" frontier --case release-demo
cg --workspace "$WORKSPACE" blockers --case release-demo
For the full walkthrough:
Global CLI install:
npm install -g @caphtech/casegraph-cli
cg --help
Published package set:
@caphtech/casegraph-core@caphtech/casegraph-cli@caphtech/casegraph-importer-markdown@caphtech/casegraph-sink-markdown@caphtech/casegraph-worker-shell@caphtech/casegraph-worker-code-agent@caphtech/casegraph-worker-local-llmRelease and verification guides:
This section is for contributors working from a clone of this repository. External users should follow the Quickstart and Install sections above.
Common commands:
pnpm build
pnpm lint
pnpm typecheck
pnpm test
pnpm test:e2e
pnpm test:analysis-golden
pnpm test:analysis-eval
The CLI entrypoint used in docs and tests is:
pnpm run cg --workspace "$WORKSPACE" ...
docs/ specs, ADRs, guides, examples
packages/ core, cli, importer, sink, workers
tests/ unit, integration, conformance, property, e2e
This repository is licensed under the Apache License 2.0. See LICENSE.
Copyright 2026 CAPH TECH Inc.
npx claudepluginhub caphtech/casegraph --plugin casegraph-pluginPersistent knowledge management with pce-memory v2: raw observe, distill/promote pipeline, intent-aware activate, and layer-aware sync
State machine for AI agent workflow management
Use FlutterHelm through the bundled npx wrapper for stable Flutter workflows and opt-in beta companion guidance.
zellij integration skills for Claude Code - worktree management with zellij tabs
OpenAI Codex (gpt-5.4) 連携プラグイン。Codex MCP サーバーを提供し、セカンドオピニオン・ペアプログラミング・構造化議論スキルを統合。
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Binary reverse engineering, malware analysis, firmware security, and software protection research for authorized security research, CTF competitions, and defensive security