By jl-cmd
Workflow enforcement for Claude Code skills. Adds a Stop hook that blocks session completion until all defined workflow steps are finished.
Workflow enforcement for Claude Code skills. A Stop hook blocks session completion until all defined workflow steps are marked done.
/plugin marketplace add jl-cmd/claude-workflow
/plugin install claude-workflow@jl-cmd-claude-workflow
~/.claude/runtime/workflow-state.json.completed array as it finishes.hooks/verify-completion.py) runs when a session tries to end. If any steps remain incomplete, it blocks completion with a message listing what is left.{
"workflow": "my-workflow",
"steps": ["step-a", "step-b", "step-c"],
"completed": ["step-a"]
}
workflow -- name shown in the blocking messagesteps -- all steps that must completecompleted -- steps finished so farTry the built-in demo:
/workflow-demo
This runs a 3-step workflow. Try ending the session before all steps finish -- the hook will block you.
---
name: my-skill
hooks:
- type: Stop
script: hooks/verify-completion.py
---
import json, os
state_path = os.path.expanduser("~/.claude/runtime/workflow-state.json")
os.makedirs(os.path.dirname(state_path), exist_ok=True)
state = {
"workflow": "my-skill",
"steps": ["step-1", "step-2", "step-3"],
"completed": [],
}
with open(state_path, "w") as f:
json.dump(state, f, indent=2)
completed array:import json
state_path = os.path.expanduser("~/.claude/runtime/workflow-state.json")
with open(state_path) as f:
state = json.load(f)
state["completed"].append("step-1")
with open(state_path, "w") as f:
json.dump(state, f, indent=2)
completed, the Stop hook allows the session to end normally.stop_hook_active: true, it exits immediately to prevent infinite loops.CLAUDE_WORKFLOW_STATE_PATH environment variable (used in tests).MIT
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.
npx claudepluginhub jl-cmd/claude-workflow --plugin workflowMemory consolidation for Claude Code. Simulates the Auto Dream feature by auditing, deduplicating, and rebuilding your memory files.
Deep research with official-docs-first methodology. Includes /deep-research (iterative multi-source research with Obsidian reports) and /research-mode (anti-hallucination constraints with source authority hierarchy).
Agentic workflow mechanics: CLAUDE.md instruction quality and structured context handoff
Capture repeatable session workflows into reusable SKILL.md files. An open-source reimplementation of the internal /skillify skill.
Forces skill evaluation before every response
Enforces task clarity before execution and requires verified work before exit
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
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).