test(qa): Add task-based example runner

package.json

@@ -11,7 +11,8 @@
     "typecheck": "pnpm -r typecheck",
     "check": "pnpm lint && pnpm typecheck && pnpm test",
     "dev": "pnpm --filter @sentry/dotagents dev",
-    "smoke:examples": "pnpm build && node scripts/smoke-examples.mjs"
+    "smoke:examples": "pnpm qa:example",
+    "qa:example": "pnpm build && node skills/dotagents-qa/scripts/qa-example.mjs"
   },
   "simple-git-hooks": {
     "pre-commit": "pnpm lint-staged"

skills/dotagents-qa/SKILL.md

@@ -34,9 +34,9 @@ Write down the QA target before running commands:
   broken state, user-scope state, or remote source.
 
 Read the targeted reference before running runtime-specific QA:
-- Core install/sync example: [references/core-smoke.md](references/core-smoke.md)
-- Codex custom agents: [references/codex.md](references/codex.md)
-- Claude Code files/runtime caveats: [references/claude.md](references/claude.md)
+- Core install/sync example: [references/core-agentic-qa.md](references/core-agentic-qa.md)
+- Codex custom agents and runtime caveats: [references/codex.md](references/codex.md)
+- Claude Code files and runtime caveats: [references/claude.md](references/claude.md)
 - Cursor files/runtime caveats: [references/cursor.md](references/cursor.md)
 - OpenCode files/runtime caveats: [references/opencode.md](references/opencode.md)
 
@@ -119,7 +119,7 @@ su -s /bin/bash node -c '
   pnpm install --frozen-lockfile
   pnpm build
   pnpm check
-  pnpm smoke:examples
+  pnpm qa:example
 '
 ```
 
@@ -128,17 +128,17 @@ the check is already known to be unrelated. If `build` or `check` fails, treat
 that as a QA finding and stop before fixture work unless you are explicitly
 isolating the playbook mechanics. If skipped or bypassed, report why.
 
-## 3. Prefer The Checked-In Smoke
+## 3. Prefer The Checked-In Agentic QA
 
-Use the checked-in example smoke for ordinary install/sync QA:
+Use the checked-in example QA for ordinary install/sync QA:
 
 ```bash
-pnpm smoke:examples
+pnpm qa:example
 ```
 
-The smoke builds the local CLI, copies `examples/full/` to a temp project, and
+The QA runner builds the local CLI, copies `examples/full/` to a temp project, and
 asserts:
-- `install`, `list`, `doctor --fix`, and `doctor`
+- `install`, `list`, `doctor --fix`, and `doctor` complete successfully
 - managed skills under `.agents/skills/`
 - Claude/Cursor skill symlink behavior
 - MCP files for Claude, Cursor, Codex, and OpenCode
@@ -147,15 +147,15 @@ asserts:
 - generated subagent runtime files for Claude, Cursor, Codex, and OpenCode
 - `sync` repair after deleting representative generated files
 
-Use `node scripts/smoke-examples.mjs --keep` when you need to inspect the temp
-project; the script prints the retained path.
+Use `node skills/dotagents-qa/scripts/qa-example.mjs all --keep` when you need
+to inspect the temp project; the script prints the retained path.
 
 For paid Codex runtime proof of generated custom agents, run the runtime proof
 outside Docker only when the branch affects Codex custom agents or when
 reporting that Codex itself works:
 
 ```bash
-node scripts/smoke-examples.mjs --codex-runtime --keep
+node skills/dotagents-qa/scripts/qa-example.mjs codex-runtime --keep
 ```
 
 That mode copies Codex auth/config into a temp `CODEX_HOME`, marks only the
@@ -320,10 +320,10 @@ opencode --version
 ```
 
 Codex subagents need real runtime proof before claiming Codex loaded them. Use
-`node scripts/smoke-examples.mjs --codex-runtime --keep`; `codex debug
-prompt-input` is not enough unless it visibly includes the generated agent name
-or instructions. Project-scoped `.codex/agents/` load only when Codex trusts
-the project. See [references/codex.md](references/codex.md).
+`node skills/dotagents-qa/scripts/qa-example.mjs codex-runtime --keep`;
+`codex debug prompt-input` is not enough unless it visibly includes the
+generated agent name or instructions. Project-scoped `.codex/agents/` load only
+when Codex trusts the project. See [references/codex.md](references/codex.md).
 
 Claude has no cheap dry-run skill list. If auth/network/model cost is
 acceptable, run a minimal non-interactive prompt from the temp project;

skills/dotagents-qa/SOURCES.md

@@ -20,7 +20,7 @@
 
 ## Decisions
 
-- Keep the runtime skill as a guided QA playbook rather than a broad QA matrix or fixed smoke harness.
+- Keep the runtime skill as a guided QA playbook rather than a broad QA matrix or fixed all-in-one harness.
 - Provide a repo-local Dockerfile for the sandbox/toolchain only; keep fixture design and assertions manual.
 - Keep the Dockerfile's prepared pnpm version aligned with the root `packageManager`.
 - Mount the host checkout read-only and do dependency install/build inside Docker to avoid host system-file writes and host binary assumptions.

skills/dotagents-qa/SPEC.md

@@ -19,7 +19,7 @@ Out of scope:
 - release publishing
 - network-backed source testing for ordinary install-location changes
 - replacing focused Vitest regression coverage for logic bugs
-- treating a fixed smoke script as sufficient QA for behavior-specific changes
+- treating a fixed all-in-one QA task as sufficient for behavior-specific changes
 
 ## Runtime Contract
 
@@ -36,6 +36,6 @@ Out of scope:
 ## Maintenance
 
 - Keep `SKILL.md` focused on guided, change-specific Docker QA.
-- Keep examples editable and illustrative; do not turn the skill into a fixed test harness.
+- Keep examples editable and illustrative; do not turn the skill into a fixed all-in-one harness.
 - Keep the Dockerfile pnpm version aligned with the root `packageManager`.
 - Update examples when dotagents changes config fields, generated file locations, or supported agents.

skills/dotagents-qa/references/claude.md

@@ -4,7 +4,7 @@ Use this reference when changes affect Claude Code skill symlinks, `.claude/sett
 
 ## File-Level Checks
 
-The core smoke asserts:
+The core agentic QA asserts:
 
 - `.claude/skills` is a symlink to `.agents/skills`
 - `.mcp.json` exists for Claude MCP config

skills/dotagents-qa/references/codex.md

@@ -4,7 +4,7 @@ Use this reference when changes affect Codex config generation, `.codex/agents/*
 
 ## File-Level Checks
 
-The core smoke asserts:
+The core agentic QA asserts:
 
 - `.codex/config.toml` exists for MCP config
 - `.codex/agents/code-reviewer.toml` exists
@@ -19,10 +19,10 @@ These checks prove dotagents wrote the expected files. They do not prove Codex l
 Run the paid runtime proof when the branch affects Codex custom agents or when reporting that Codex itself works:
 
 ```bash
-node scripts/smoke-examples.mjs --codex-runtime --keep
+node skills/dotagents-qa/scripts/qa-example.mjs codex-runtime --keep
 ```
 
-The script:
+The task:
 
 - copies `examples/full/` to a temp project
 - runs the built local dotagents CLI
@@ -48,4 +48,4 @@ Project-scoped `.codex/` layers load only when Codex trusts the project. A one-o
 
 If Codex reports `unknown agent_type`, check project trust and the canonical path (`pwd -P`) before assuming the generated TOML schema is wrong.
 
-Do not leave copied Codex auth in retained temp directories. The smoke script scrubs its temp `codex-home`; if you run manual experiments, remove copied auth/config before reporting.
+Do not leave copied Codex auth in retained temp directories. The QA task scrubs its temp `codex-home`; if you run manual experiments, remove copied auth/config before reporting.

skills/dotagents-qa/references/core-smoke.md → skills/dotagents-qa/references/core-agentic-qa.md

@@ -1,4 +1,4 @@
-# Core Smoke
+# Core Agentic QA
 
 Use this reference for ordinary dotagents install/sync QA before drilling into a specific runtime.
 
@@ -7,12 +7,12 @@ Use this reference for ordinary dotagents install/sync QA before drilling into a
 Run:
 
 ```bash
-pnpm smoke:examples
+pnpm qa:example
 ```
 
 This builds the local CLI, copies `examples/full/` to a temp project, and verifies:
 
-- `install`, `list`, `doctor --fix`, and `doctor`
+- `install`, `list`, `doctor --fix`, and `doctor` complete successfully
 - managed skills under `.agents/skills/`
 - Claude/Cursor skill symlink behavior
 - MCP files for Claude, Cursor, Codex, and OpenCode
@@ -21,7 +21,8 @@ This builds the local CLI, copies `examples/full/` to a temp project, and verifi
 - generated subagent runtime files for Claude, Cursor, Codex, and OpenCode
 - `sync` repair after deleting representative generated files
 
-Use `node scripts/smoke-examples.mjs --keep` to keep the temp project for inspection. The script prints the project path.
+Use `node skills/dotagents-qa/scripts/qa-example.mjs all --keep` to keep the
+temp project for inspection. The script prints the project path.
 
 ## What This Proves
 

skills/dotagents-qa/references/cursor.md

@@ -4,7 +4,7 @@ Use this reference when changes affect Cursor skill sharing, `.cursor/mcp.json`,
 
 ## File-Level Checks
 
-The core smoke asserts:
+The core agentic QA asserts:
 
 - Cursor shares Claude-compatible skills through `.claude/skills`
 - `.cursor/mcp.json` exists for Cursor MCP config

skills/dotagents-qa/references/opencode.md

@@ -4,7 +4,7 @@ Use this reference when changes affect OpenCode config generation, `opencode.jso
 
 ## File-Level Checks
 
-The core smoke asserts:
+The core agentic QA asserts:
 
 - `opencode.json` exists for OpenCode MCP config
 - `.opencode/agents/code-reviewer.md` exists