Paths: File paths (shared/, references/, ../ln-*) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root. If shared/ is missing, fetch files via WebFetch from https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path}.

Dev Environment Setup

Type: L2 Domain Coordinator Category: 0XX Shared

Assess-Dispatch-Verify coordinator. Probes entire environment once, builds dispatch plan, invokes only workers that have work to do, then verifies everything. Runs all phases in one uninterrupted pass.

When to Use This Skill

• First-time project setup (no docs/environment_state.json)
• After installing/removing CLI agents (Codex, Gemini)
• After adding/removing MCP servers in Claude Code
• When hooks need reconfiguration or verification
• Periodic alignment check across all agents

Input Parameters

Workflow

Detect → Assess → Dispatch → Verify & Report

Phase 0: Environment Detection

Config paths by OS:

Load disabled flags: Read docs/environment_state.json if exists. Extract agents.{name}.disabled for each agent.

Phase 1: ASSESS (all probes, one pass, read-only)

Single pass that collects the full environment state. No mutations — read-only probes.

1a: CLI Agent Probes

For each non-disabled agent, run {agent} --version:

Skip disabled agents. Record "not found" for missing agents.

1b: MCP Server Probes

One claude mcp list call. Parse into map per server: registered, connected.

Then for each registered hex package: npm outdated -g @levnikolaevich/{pkg} — captures outdated: true/false.

Detect deprecated servers (from ln-012 deprecated list in its SKILL.md).

1c: Hooks & Permissions Probe

Read ~/.claude/settings.json (or settings.local.json):

• Hooks present? (PreToolUse, PostToolUse, SessionStart)
• disableAllHooks flag?
• permissions.allow[] — which mcp__* entries present?

MANDATORY READ: Load shared/references/hook_health_check.md — run script existence + dependency checks.

1d: Config Sync Probe

For each non-disabled target agent:

• Check symlink status (~/.gemini/skills, ~/.codex/skills)
• Read target config — compare MCP servers against Claude source
• Check hook mappings (Gemini only)

1e: Instruction Files Probe

Check existence of: CLAUDE.md, AGENTS.md, GEMINI.md in project root.

For each existing file: line count, estimated tokens, has timestamps, has compact instructions, has MCP Tool Preferences table.

Assessment Summary Table:

Environment Assessment:
| Area               | Status | Detail                              |
|--------------------|--------|-------------------------------------|
| CLI Agents         | 2/3 ok | codex: not found                    |
| MCP Servers        | 5 conn | hex-line outdated (1.3.5 → 1.3.6)  |
| Hooks              | ok     | 3/3 configured                      |
| Permissions        | 4/5    | mcp__hex-ssh missing                |
| Config Sync        | gemini | codex: symlink missing              |
| Instruction Files  | 1/3    | AGENTS.md, GEMINI.md missing        |

Phase 2: DISPATCH (selective invocation)

2a: Decision Matrix

2b: Dispatch Plan

Display the plan before invoking:

Dispatch Plan:
| Worker | Action | Reason                           |
|--------|--------|----------------------------------|
| ln-011 | RUN    | codex not installed               |
| ln-012 | SKIP   | all servers connected, hooks ok   |
| ln-013 | RUN    | codex symlink missing             |
| ln-014 | RUN    | AGENTS.md, GEMINI.md missing      |

2c: Fast Path — if ALL workers SKIP → go directly to Phase 3 with message: "All green — no workers needed."

2d: Invoke (sequential, no stops between invocations):

For each worker where action == RUN:
  Skill(skill: "{worker}", args: "{OS} {disabled_flags} {dry_run}")

Workers check their own zone and fix what's needed. ln-010 does NOT tell them what to do — only decides whether to call them.

Phase 3: VERIFY & REPORT

Full verification after workers complete. This is the acceptance check — confirms everything is correct.

3a: Verify CLI Agents

• codex --version, gemini --version, claude --version for each non-disabled agent
• Record: available, version

3b: Verify MCP Servers

• claude mcp list → verify all registered servers Connected
• Token budget: server_count * 5000 tokens (% of 200K context)

3c: Verify Hooks

MANDATORY READ: Load shared/references/hook_health_check.md

• Validate hooks from hooks/hooks.json (plugin) and .claude/settings.local.json (project)
• Check JSON syntax, script existence, dependencies

3d: Verify Instruction Files

• Exists in project root
• Line count (<100 recommended)
• No timestamps (breaks prompt cache)
• Has compact instructions section

3e: Best Practices Audit

3f: Write State

1. Read existing docs/environment_state.json (preserve disabled flags)
2. Build new state validated against references/environment_state_schema.json:
   • scanned_at: ISO 8601 timestamp
   • agents: probe results merged with preserved disabled, plus config sync data per agent
   • claude_md: exists, line_count, has_timestamps, has_compact_instructions
   • best_practices: score and findings
   • last_assessment: dispatch plan outcome (workers_run, workers_skipped)
   • NOT persisted (verify-time only): MCP servers, MCP budget, hooks — shown in report but not written to file (native CC settings are source of truth)
3. Ensure docs/ directory exists
4. Ensure gitignore: If inside a git repo, check .gitignore for docs/environment_state.json. If not covered → append:
   Copy

   # Machine-specific environment state (generated by ln-010)
   docs/environment_state.json
   

5. Write docs/environment_state.json (2-space indent)

3g: Summary Report

Dev Environment Setup Complete:
| Area              | Result   | Detail                        |
|-------------------|----------|-------------------------------|
| CLI agents        | ok       | codex installed v0.1.2503     |
| MCP servers       | ok       | 5 configured, budget OK       |
| Config sync       | ok       | codex: symlink + 5 servers    |
| Hooks             | ok       | 3/3 active                    |
| Instructions      | ok       | AGENTS.md, GEMINI.md created  |
| Best practices    | 9/9      | All green                     |

Dispatch: 3 workers run (ln-011, ln-013, ln-014), 1 skipped (ln-012)
State: docs/environment_state.json

If ANY problems found → list them explicitly after the table with suggested fix per issue.

Worker Invocation (MANDATORY)

All workers: Invoke via Skill tool. Do NOT stop or pause between invocations — execute the entire pipeline from Phase 0 to Phase 3 in a single uninterrupted pass.

TodoWrite format (mandatory):

- Detect environment (pending)
- Assess: probe all areas (pending)
- Dispatch: build plan + invoke workers (pending)
- Verify & report (pending)

Critical Rules

Anti-Patterns

Meta-Analysis

MANDATORY READ: Load shared/references/meta_analysis_protocol.md

Skill type: execution-orchestrator. Analyze this session per protocol §7. Output per protocol format.

Definition of Done

• OS and environment detected (Phase 0)
• Full assessment completed — all areas probed (Phase 1)
• Dispatch plan built and displayed (Phase 2)
• Only needed workers invoked (Phase 2d)
• Full verification pass completed (Phase 3)
• Best practices audit table shown with 9 checks (Phase 3e)
• docs/environment_state.json written with dispatch plan outcome (Phase 3f)
• docs/environment_state.json covered in .gitignore
• Existing disabled flags preserved across rescan
• Summary report displayed with dispatch outcome and problems listed if any (Phase 3g)

Version: 2.0.0 Last Updated: 2026-03-23