From tenequm-skills
OpenClaw platform reference - plugin system, extensions, configuration, boot/provisioning, channels, models, CLI. Use when working on openclaw codebase, building openclaw plugins/extensions, configuring openclaw instances, provisioning openclaw gateways, designing agent provisioning flows (e.g. agentbox), or debugging openclaw config/plugin/channel issues. Triggers on openclaw, openclaw config, openclaw plugin, openclaw extension, openclaw channel, openclaw gateway, openclaw provisioning, openclaw onboarding, openclaw boot, openclaw skills, BOOT.md, openclaw.plugin.json, openclaw-x402, agentbox provisioning.
npx claudepluginhub tenequm/skills --plugin gh-cliThis skill uses the workspace's default tool permissions.
Structured reference for the OpenClaw assistant platform. Source repo: `~/Projects/openclaw/`. Tracking version: **2026.4.7** (refresh: 2026-04-07).
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.
Checks Next.js compilation errors using a running Turbopack dev server after code edits. Fixes actionable issues before reporting complete. Replaces `next build`.
Structured reference for the OpenClaw assistant platform. Source repo: ~/Projects/openclaw/. Tracking version: 2026.4.7 (refresh: 2026-04-07).
Load the relevant reference file based on the task:
references/plugin-system.md - Plugin discovery, loading, manifest format, registration API, installation CLI, SDK exports. Read when building/debugging plugins or understanding how extensions are loaded.references/configuration.md - Full config structure (OpenClawConfig), config paths, file format (JSON5), env var substitution, $include directives, validation (Zod), per-plugin config, programmatic read/write. Read when modifying or generating openclaw config.references/boot-provisioning.md - Gateway startup sequence, BOOT.md mechanism, onboarding flows (interactive/non-interactive), plugin loading during boot, sidecar startup order, health checks. Read when designing automated provisioning or debugging startup.references/channels-extensions.md - All built-in channels (Telegram, Discord, Slack, Signal, iMessage, WhatsApp, Web) + extension channels (MS Teams, Matrix, Zalo, Voice Call, Feishu). Channel plugin registration, per-channel config, Telegram specifics. Read when adding/configuring channels.references/models-providers.md - Provider configuration, models.mode (merge/replace), x402 providers, model catalog structure, provider registration via plugins. Read when configuring model access or building provider plugins.references/cli-commands.md - Key CLI commands for config, plugins, channels, agents, onboarding, gateway, skills. Read when scripting openclaw setup or building automation.references/github-context.md - Open bugs, breaking changes, recent impactful PRs, plugin/config known issues, dev gotchas synthesized from GitHub. Refreshed via /refresh-openclaw. Read before starting any non-trivial openclaw work to avoid known pitfalls.src/plugins/ - Plugin loader, discovery, registry, install, types
src/plugin-sdk/ - 100+ scoped exports for plugin consumption
src/config/ - Config loading, types, validation, defaults, paths
src/gateway/ - Gateway server, boot, startup, server methods
src/commands/ - CLI commands (onboard, config, etc.)
src/cli/ - CLI wiring (plugins-cli, skills-cli, etc.)
extensions/ - Extension implementations (channels, memory, etc.)
skills/ - Built-in skill definitions
docs/ - Documentation (Mintlify)
~/.openclaw/openclaw.json (parsed as JSON5)~/.openclaw/extensions/~/.openclaw/skills/ (managed) + workspace skills~/.openclaw/agents/<agent-id>/~/.openclaw/agents/<agent-id>/sessions/~/.openclaw/credentials/OPENCLAW_CONFIG_PATH env var{
"id": "plugin-id",
"name": "Display Name",
"kind": "memory",
"channels": ["channel-id"],
"providers": ["provider-id"],
"configSchema": { "type": "object" },
"uiHints": { "field": { "label": "...", "sensitive": true } }
}
Top-level keys in OpenClawConfig:
meta, auth, acp, env, secrets, plugins, skills, models, agents, tools, channels, session, hooks, gateway, logging, browser, memory, messages, approvals, cron