feat(agents): add native plugins for Codex Claude and Cursor

.agents/plugins/marketplace.json

@@ -0,0 +1,20 @@
+{
+  "name": "hack-dance",
+  "interface": {
+    "displayName": "Hack Dance"
+  },
+  "plugins": [
+    {
+      "name": "hack",
+      "source": {
+        "source": "local",
+        "path": "./plugins/hack"
+      },
+      "policy": {
+        "installation": "AVAILABLE",
+        "authentication": "ON_INSTALL"
+      },
+      "category": "Developer Tools"
+    }
+  ]
+}

.claude-plugin/marketplace.json

@@ -0,0 +1,16 @@
+{
+  "name": "hack-dance",
+  "description": "Official Hack plugins for local-first development workflows.",
+  "owner": {
+    "name": "Hack Dance LLC"
+  },
+  "plugins": [
+    {
+      "name": "hack",
+      "source": "./plugins/hack",
+      "description": "Operate and onboard Hack-managed local development environments.",
+      "category": "development",
+      "tags": ["local-development", "docker", "developer-tools", "agents"]
+    }
+  ]
+}

.claude/settings.json

@@ -11,12 +11,5 @@
         ]
       }
     ]
-  },
-  "mcpServers": {
-    "hack": {
-      "type": "stdio",
-      "command": "hack",
-      "args": ["mcp", "serve"]
-    }
   }
 }

.cursor-plugin/marketplace.json

@@ -0,0 +1,16 @@
+{
+  "name": "hack-dance",
+  "owner": {
+    "name": "Hack Dance LLC"
+  },
+  "metadata": {
+    "description": "Official Hack plugins for local-first development workflows."
+  },
+  "plugins": [
+    {
+      "name": "hack",
+      "source": "plugins/hack",
+      "description": "Operate and onboard Hack-managed local development environments."
+    }
+  ]
+}

.releaserc.json

@@ -24,7 +24,10 @@
           "packages/db/package.json",
           "services/auth-broker/package.json",
           "bun.lock",
-          "apps/macos/Config/Base.xcconfig"
+          "apps/macos/Config/Base.xcconfig",
+          "plugins/hack/.codex-plugin/plugin.json",
+          "plugins/hack/.claude-plugin/plugin.json",
+          "plugins/hack/.cursor-plugin/plugin.json"
         ],
         "message": "chore(release): ${nextRelease.version}\n\n${nextRelease.notes}"
       }

AGENTS.md

@@ -210,11 +210,11 @@ Most formatting and common issues are automatically fixed by Biome. Run `bun x u
 Use `hack` as the single interface for local-first runtime orchestration (compose, DNS/TLS, logs, env, and persistent project workspaces).
 
 Integration freshness:
-- These instructions were generated by hack CLI v3.5.0; treat cached rules from another version as potentially stale.
+- These instructions were generated by hack CLI v3.5.1; treat cached rules from another version as potentially stale.
 - At session start, audit project and global integrations with `hack setup sync --all-scopes --check`.
 - If anything is stale, missing, or deprecated, run `hack setup sync --all-scopes`, then reload the agent session so cached instructions are replaced.
 - Never copy or hand-edit generated Hack rules to refresh them; update the CLI and run the sync command.
-- Content revision: `1319d93c89c3` (version alone is not a freshness guarantee).
+- Content revision: `6ef36575c6e7` (version alone is not a freshness guarantee).
 
 Product boundary:
 - Supported v3 surface: project init, up/down/restart, open, logs, env, host exec/shell, sessions, doctor, and daemon.
@@ -366,7 +366,7 @@ Docker compose notes:
 - Use `docker compose -f .hack/docker-compose.yml exec <service> <cmd>` only if you need exec into a running container.
 
 Agent integration maintenance:
-- Project-level hack commands auto-check integration drift and attempt auto-sync (project docs, client skills/rules, shared global skills, and MCP).
+- Project-level hack commands auto-check integration drift and attempt auto-sync (project docs, Cursor/Claude integrations, shared global skills, and MCP).
 - When drift is detected, Hack reports it before repair and tells the agent to reload after repair; it never silently leaves the session using cached rules.
 - Set `HACK_SETUP_SYNC_MODE=warn` to only warn, or `HACK_SETUP_SYNC_MODE=off` to disable.
 - Refresh project + user integrations: `hack setup sync --all-scopes`
@@ -376,9 +376,9 @@ Agent integration maintenance:
 - When changing hack itself: interface or behavior changes must update docs/ in the same change (regenerate the CLI reference with `bun run docs:cli-reference`).
 
 Agent setup (CLI-first):
-- Cursor rules: `hack setup cursor`
-- Claude hooks: `hack setup claude`
-- Codex skill: `hack setup codex`
+- Cursor plugin: run `hack setup cursor`, then install Hack with `/add-plugin` or enable it from Settings > Plugins as directed.
+- Claude Code plugin: run `hack setup claude`, then install or enable `hack@hack-dance` as directed.
+- Codex plugin: run `hack setup codex`, then install or enable Hack from `/plugins` as directed.
 - Refresh all local agent integrations: `hack setup sync --all-scopes`
 - Agent-assisted onboarding: `hack init --with claude|codex|both` (new repos) or `hack agent onboard` (existing projects) print/hand off the full setup prompt; the `/hack-init` skill and the `hack-init` MCP prompt return the same content.
 - Init prompt: `hack agent init` (use --client cursor|claude|codex to open)

CLAUDE.md

@@ -80,11 +80,11 @@ This project uses Obsidian for project context, specs, research, and progress tr
 Use `hack` as the single interface for local-first runtime orchestration (compose, DNS/TLS, logs, env, and persistent project workspaces).
 
 Integration freshness:
-- These instructions were generated by hack CLI v3.5.0; treat cached rules from another version as potentially stale.
+- These instructions were generated by hack CLI v3.5.1; treat cached rules from another version as potentially stale.
 - At session start, audit project and global integrations with `hack setup sync --all-scopes --check`.
 - If anything is stale, missing, or deprecated, run `hack setup sync --all-scopes`, then reload the agent session so cached instructions are replaced.
 - Never copy or hand-edit generated Hack rules to refresh them; update the CLI and run the sync command.
-- Content revision: `1319d93c89c3` (version alone is not a freshness guarantee).
+- Content revision: `6ef36575c6e7` (version alone is not a freshness guarantee).
 
 Product boundary:
 - Supported v3 surface: project init, up/down/restart, open, logs, env, host exec/shell, sessions, doctor, and daemon.
@@ -236,7 +236,7 @@ Docker compose notes:
 - Use `docker compose -f .hack/docker-compose.yml exec <service> <cmd>` only if you need exec into a running container.
 
 Agent integration maintenance:
-- Project-level hack commands auto-check integration drift and attempt auto-sync (project docs, client skills/rules, shared global skills, and MCP).
+- Project-level hack commands auto-check integration drift and attempt auto-sync (project docs, Cursor/Claude integrations, shared global skills, and MCP).
 - When drift is detected, Hack reports it before repair and tells the agent to reload after repair; it never silently leaves the session using cached rules.
 - Set `HACK_SETUP_SYNC_MODE=warn` to only warn, or `HACK_SETUP_SYNC_MODE=off` to disable.
 - Refresh project + user integrations: `hack setup sync --all-scopes`
@@ -246,9 +246,9 @@ Agent integration maintenance:
 - When changing hack itself: interface or behavior changes must update docs/ in the same change (regenerate the CLI reference with `bun run docs:cli-reference`).
 
 Agent setup (CLI-first):
-- Cursor rules: `hack setup cursor`
-- Claude hooks: `hack setup claude`
-- Codex skill: `hack setup codex`
+- Cursor plugin: run `hack setup cursor`, then install Hack with `/add-plugin` or enable it from Settings > Plugins as directed.
+- Claude Code plugin: run `hack setup claude`, then install or enable `hack@hack-dance` as directed.
+- Codex plugin: run `hack setup codex`, then install or enable Hack from `/plugins` as directed.
 - Refresh all local agent integrations: `hack setup sync --all-scopes`
 - Agent-assisted onboarding: `hack init --with claude|codex|both` (new repos) or `hack agent onboard` (existing projects) print/hand off the full setup prompt; the `/hack-init` skill and the `hack-init` MCP prompt return the same content.
 - Init prompt: `hack agent init` (use --client cursor|claude|codex to open)

README.md

@@ -46,7 +46,8 @@ the branch — review a PR while your main checkout keeps running.
 **Built for agents.** `hack init --with claude|codex` hands setup to an agent.
 Machine surfaces are first-class: `--json` envelopes with stable error codes,
 `--no-interactive`, and agent instructions that sync themselves into
-AGENTS.md/CLAUDE.md, Cursor rules, and Codex skills.
+AGENTS.md/CLAUDE.md. Cursor, Claude Code, and Codex support ship as native,
+user-installed Hack plugins from this repository's marketplaces.
 
 ## More than ports
 

docs/README.md

@@ -59,7 +59,7 @@ This section is easy to find, but it does not lead the product story.
 - Working notes and historical design docs live in `docs/plans/`; they are not the supported product contract
 - Root scripts orchestrate workspace tasks through Turbo
 - Package-local commands remain available via `bun run --cwd <workspace> <script>`
-- Agent-facing docs and setup output are generated from `src/agents/instruction-source.ts` via `src/agents/*` and `src/mcp/agent-docs.ts`; update the canonical source (not the rendered surfaces) when changing AGENTS/Codex/MCP guidance.
+- Agent-facing docs, bundled Codex/Claude/Cursor plugin content, and setup output are generated from `src/agents/instruction-source.ts` via `src/agents/*` and `src/mcp/agent-docs.ts`; update the canonical source (not the rendered surfaces) when changing agent guidance.
 - The generated [CLI reference](reference/cli.md) is rendered from the command spec (`bun run docs:cli-reference`); a drift test fails when it is stale.
 - End-to-end coverage that backs docs claims lives in `tests/e2e/` (`bun run test:e2e:local`).
 - `.factory/validation/` contains historical evidence from retired hosted/web/integration work and is not the current product contract.

docs/cli.md

@@ -23,7 +23,8 @@ terminal).
 - `hack doctor` / `hack doctor --fix` — validate and repair local setup
 - `hack daemon` — optional local daemon for faster JSON status/ps
 - `hack agent onboard` — agent-assisted onboarding for existing projects
-- `hack setup` — install/refresh agent integrations (Cursor rules, Claude hooks, Codex skill, MCP)
+- `hack setup` — check/prepare native Cursor, Claude Code, and Codex plugins; refresh project
+  instructions; or explicitly configure standalone MCP
 - `hack tickets` — deprecated compatibility surface for existing Tickets data
 
 Interactive diagnostics use compact status rows: healthy groups stay on one line, while warnings
@@ -66,8 +67,9 @@ See [Beta workflows](beta.md) for guides on this surface.
   apply documented defaults or fail fast with `E_INTERACTIVE_REQUIRED`.
 - `NO_COLOR` (or `HACK_NO_COLOR`) disables colored/decorated output.
 
-Generated agent docs, Cursor rules, Codex skills, and the shared `~/.ai/skills/hack-cli` skill carry
-the Hack CLI version that generated them. Audit both project and global surfaces with
+Generated agent docs, native plugin content, and the shared `~/.ai/skills/hack-cli` skill carry the Hack CLI
+version that generated them. Native agent plugins are versioned with the Hack release and installed once
+from the `hack-dance/hack` marketplace. Audit project and global generated surfaces with
 `hack setup sync --all-scopes --check`; repair them with `hack setup sync --all-scopes`, then reload
 the agent session so it stops using cached guidance. Interactive project commands also report drift
 before auto-repair instead of repairing silently.

docs/docs-ia.md

@@ -21,8 +21,9 @@ remote/control-plane features.
 - `docs/guides/global-settings.md` — global config
 - `docs/guides/agent-first-setup.md` — agent-assisted onboarding (`hack init --with`, `hack agent onboard`)
 - `docs/guides/prerequisite-detection-matrix.md` — environment prerequisite checks
-- Generated agent-facing guidance: `AGENTS.md`/`CLAUDE.md` snippets, `.codex/skills/hack-cli/SKILL.md`,
-  and `hack-init` skill/MCP prompt content, all rendered from `src/agents/instruction-source.ts`
+- Generated agent-facing guidance: `AGENTS.md`/`CLAUDE.md` snippets, the Hack plugin's bundled
+  `hack-cli` skill, and `hack-init` skill/MCP prompt content, all rendered from
+  `src/agents/instruction-source.ts`
   (canonical source) via `src/agents/*` and `src/mcp/agent-docs.ts`.
 
 New docs belong here when they describe supported v3 CLI surface a user hits without opting into
@@ -92,7 +93,7 @@ extend `hack` (authoring extensions, SDK usage) rather than how to use it.
 
 `docs/reference/cli.md` is generated from the command specs (`bun run docs:cli-reference`) and
 must not be hand-edited. Agent-facing instruction surfaces (`AGENTS.md`/`CLAUDE.md` snippets, the
-Codex skill, Cursor rules, the session primer) are all rendered from the canonical section list in
+native plugin skills/rules, the session primer) are all rendered from the canonical section list in
 `src/agents/instruction-source.ts`; update that file, not the individual renderers, when agent
 guidance needs to change, then regenerate.
 

docs/guides/agent-first-setup.md

@@ -11,7 +11,7 @@ your installed version.
 | --- | --- |
 | A fresh repo + an agent CLI installed | `hack init --with claude\|codex\|both` |
 | An existing hack project | `hack agent onboard` (prints the prompt) |
-| An agent session with the hack skills installed | `/hack-init` |
+| A Codex session with the Hack plugin installed | `/hack-init` |
 | A no-shell MCP client | the `hack-init` MCP prompt (`hack setup mcp`) |
 
 ## New repo: `hack init --with`
@@ -42,11 +42,11 @@ when present.
 
 ## Agent-side skill: `/hack-init`
 
-`hack setup claude` and `hack setup codex` install a thin `hack-init` skill
-(`.claude/skills/hack-init/SKILL.md` / `.codex/skills/hack-init/SKILL.md`).
-The skill tells the agent to run `hack agent onboard` (or fetch the `hack-init`
-MCP prompt) and follow it — the content stays in the CLI, so installed skills
-never go stale on substance.
+The official Hack plugins for Cursor, Claude Code, and Codex bundle the same `hack-init` skill;
+run `hack setup cursor`, `hack setup claude`, or `hack setup codex` for exact marketplace guidance.
+The skill tells the agent to run `hack agent onboard` (or fetch the
+`hack-init` MCP prompt) and follow it — the content stays in the CLI, so installed skills never go
+stale on substance.
 
 ## What the prompt covers