ln-013-config-syncer
Use when installing or verifying this marketplace in Claude and Codex, aligning selected plugins, MCP settings, and Codex execution defaults.
What this skill does
> **Paths:** File paths (`references/`) are relative to this skill directory.
# Marketplace and Config Aligner
**Type:** L3 Worker
**Category:** 0XX Shared
Installs or verifies this repository's marketplace and selected plugins for Claude Code and Codex CLI without making either agent the source of truth for the other. `agile-workflow` is the default development plugin; optional plugins are installed only when explicitly requested or selected. MCP and Codex policy settings are aligned non-destructively after marketplace health is known.
## MANDATORY READ
**MANDATORY READ:** Load `references/coordinator_summary_contract.md`, `references/environment_worker_runtime_contract.md`, and `references/worker_runtime_contract.md`
**MANDATORY READ:** Load `references/agent_skill_roots_contract.md`
---
## Input / Output
| Direction | Content |
|-----------|---------|
| **Input** | OS info, `disabled` flags per agent, `targets` (`claude` / `codex` / `both` / `all`), `plugins` (`agile-workflow` default, explicit list, or `all`), `dry_run` flag, optional `auto_install_providers` flag, optional `runId`, optional `summaryArtifactPath` |
| **Output** | Structured summary envelope with `payload.status` = `completed` / `skipped` / `error`, plus per-target alignment outcomes in `changes` / `detail` |
If `summaryArtifactPath` is provided, write the same summary JSON there. If not provided, return the summary inline and remain fully standalone. If `runId` is not provided, generate a standalone `run_id` before emitting the summary envelope.
`dry_run=true` is detection-only: do not register marketplaces, install plugins, edit MCP/Codex config, write `.bak` files, or auto-install providers. Return planned actions and drift details in the summary.
## Runtime
Runtime family: `environment-worker-runtime`
Phase profile:
1. `PHASE_0_CONFIG`
2. `PHASE_1_DISCOVER_STATE`
3. `PHASE_2_VERIFY_MARKETPLACE_INSTALLS`
4. `PHASE_3_ALIGN_MCP_SETTINGS`
5. `PHASE_4_ALIGN_CODEX_POLICY`
5a. `PHASE_4A_MCP_PROVIDER_CHECK`
6. `PHASE_5_WRITE_SUMMARY`
7. `PHASE_6_SELF_CHECK`
Runtime rules:
- emit `summary_kind=env-marketplace-align`
- standalone runs generate their own `run_id` and write the default worker-family artifact path
- managed runs require both `runId` and `summaryArtifactPath` and must write the summary to the exact provided path
- always write the validated summary artifact before terminal outcome
## Output Contract
Always build a structured `env-marketplace-align` summary envelope per:
- `references/coordinator_summary_contract.md`
- `references/environment_worker_runtime_contract.md`
Payload fields:
- `targets`
- `plugin_alignment`
- `marketplace_health`
- `mcp_alignment`
- `mcp_providers`
- `codex_execution_defaults`
- `status`
---
## Config Paths by OS
| Agent | Windows | macOS / Linux |
|-------|---------|---------------|
| **Claude** (primary) | `%USERPROFILE%\.claude.json` | `~/.claude.json` |
| **Claude** (fallback) | `%USERPROFILE%\.claude\settings.json` | `~/.claude/settings.json` |
| **Codex** | `%USERPROFILE%\.codex\config.toml` | `~/.codex/config.toml` |
---
## Workflow
```text
Discover State -> Verify Marketplace Installs -> Align MCP -> Align Codex Policy -> Verify & Report
```
### Phase 1: Discover State
1. Read Claude settings:
- `~/.claude.json` (primary) + `~/.claude/settings.json` (fallback)
- extract `mcpServers`, enabled marketplaces/plugins, and hook state
2. Read Codex config:
- `~/.codex/config.toml`
- extract `[mcp_servers.*]`, `[marketplaces.*]`, top-level `approval_policy`, and top-level `sandbox_mode`
3. Inspect repo marketplace surfaces:
- Claude: `.claude-plugin/marketplace.json`
- Codex: `.agents/plugins/marketplace.json`
- Codex plugin manifests: `plugins/*/.codex-plugin/plugin.json`
4. Resolve plugin selection:
- always include `agile-workflow`
- if `plugins=all`, include every plugin listed by the marketplace
- if an explicit list is provided, install `agile-workflow` plus the requested plugins
- if interactive and optional plugins are not specified, ask which optional plugins to install
- if non-interactive and optional plugins are not specified, install or verify only `agile-workflow`
5. Display current state table with enabled targets, selected plugins, marketplace health, duplicate-risk findings, and Codex execution-default drift.
### Phase 2: Verify Marketplace Installs
For each target where `disabled` is not `true`:
| Target | Expected install model | Verification |
|--------|------------------------|--------------|
| Claude | Native Claude marketplace | Marketplace is registered through Claude plugin settings and selected plugins resolve from `.claude-plugin/marketplace.json` |
| Codex | Native Codex marketplace | Marketplace is registered through Codex plugin settings and selected plugins resolve from `.agents/plugins/marketplace.json` / `plugins/*/.codex-plugin/plugin.json` |
Install policy:
- Claude marketplace: `/plugin marketplace add levnikolaevich/claude-code-skills`
- Claude plugin: `/plugin install {plugin}@levnikolaevich-skills-marketplace`
- Codex marketplace: `codex plugin marketplace add levnikolaevich/claude-code-skills`
- Codex plugin selection: verify selected plugins exist in `.agents/plugins/marketplace.json` and `plugins/{plugin}/.codex-plugin/plugin.json`; if the installed Codex version exposes plugin-level selection only through an interactive native flow, report that exact step instead of fabricating a non-existent CLI command
Rules:
- If `dry_run=true`, do not run marketplace add/install commands; report the exact commands that would run.
- Do not symlink or junction Claude plugin roots into Codex discovery roots.
- Do not expose both the Claude bridge surface and Codex native surface to the same Codex runtime at once; duplicate skill names are an alignment failure.
- Do not delete user plugin roots. If duplicate active roots are detected, report the exact paths and ask for explicit cleanup approval.
- Do not install optional plugins silently.
- Run `node tools/marketplace/validate.mjs` when the repo root is available. Report failure as marketplace drift.
### Phase 3: Align MCP Settings
MCP settings are aligned by server name. No target is the universal source of truth.
Merge strategy:
1. Build a union of Claude `mcpServers` and Codex `[mcp_servers.*]`.
2. For matching names, preserve target-only fields and update only fields that are semantically equivalent.
3. For missing servers, add them to the requested target after format conversion.
4. Create `.bak` before modifying any config file.
5. Preserve target-only servers and unrelated config sections.
6. If `dry_run=true`, report the merge plan only; do not create `.bak` files or edit config.
Field mapping:
| Semantic field | Claude JSON | Codex TOML |
|---|---|---|
| command | `command` | `command` |
| args | `args` | `args` |
| env | `env` | `[mcp_servers.{name}.env]` |
| HTTP URL | `type: "http"` + `url` | `url` |
| headers | `headers` | `http_headers` |
Codex-only fields to preserve:
`bearer_token_env_var`, `enabled_tools`, `disabled_tools`, `startup_timeout_sec`, `tool_timeout_sec`, `enabled`, `required`
**Windows implementation note:** Config format conversions with regex or backslash escaping MUST use a temporary `.mjs` script file, not inline `node -e` or bash heredocs.
### Phase 4: Align Hooks and Codex Policy
**Claude hooks:**
- Claude hook/style sync remains Claude-only.
- Do not project Claude hooks into Codex (the wire formats differ).
**Codex hooks:**
- Do not project Claude hooks into Codex (the wire formats differ).
- Do not install product-specific Codex hook shims from this marketplace. Runtime products that need Codex hooks own their own hook scripts, trust model, verification, and service lifecycle.
- Do not write the legacy `[features] codex_hooks = true` alias. Current Codex uses the stable hooks feature and the canonical block is pure `[hooks.*]` TOML.
- When removing stale hRelated in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.