@xevy/skills 0.1.21 → 0.1.22

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -2,7 +2,7 @@
2
2
  "name": "xevy",
3
3
  "displayName": "Xevy",
4
4
  "description": "Native /xevy:* SDLC workflow commands driven through the Xevy MCP server. Cloud (server Web Agent) by default; '-here' variants run locally in this Claude Code session. Claude reviews and corrects the result after every action.",
5
- "version": "0.1.21",
5
+ "version": "0.1.22",
6
6
  "author": {
7
7
  "name": "Xevy"
8
8
  },
package/README.md CHANGED
@@ -1,6 +1,6 @@
1
1
  # @xevy/skills
2
2
 
3
- Install **Xevy SDLC agent skills** and the **Xevy MCP server** for Cursor, Claude Code, and [50+ other agents](https://github.com/vercel-labs/skills#supported-agents) with one command.
3
+ Install **Xevy SDLC agent skills** and the **Xevy MCP server** for Cursor, Claude Code, Codex, and [50+ other agents](https://github.com/vercel-labs/skills#supported-agents) with one command.
4
4
 
5
5
  This package wraps:
6
6
 
@@ -16,7 +16,7 @@ Installs are **global by default** — skills land in user scope (`~/.agents/ski
16
16
  Run from anywhere:
17
17
 
18
18
  ```bash
19
- npx @xevy/skills # global by default
19
+ npx @xevy/skills # global; defaults to Cursor + Claude Code + Codex
20
20
  ```
21
21
 
22
22
  Install into the **current project/repo** instead (writes `.cursor/skills/`, `.mcp.json`, etc. into the repo):
@@ -32,10 +32,11 @@ npx @xevy/skills -y # global by default
32
32
  npx @xevy/skills --project -y # into this repo
33
33
  ```
34
34
 
35
- Cursor only:
35
+ Target one or more specific coding agents (any identifier supported by `skills` and `add-mcp`):
36
36
 
37
37
  ```bash
38
38
  npx @xevy/skills -a cursor -y
39
+ npx @xevy/skills -a codex -a windsurf -y
39
40
  ```
40
41
 
41
42
  Custom host (self-hosted Xevy):
@@ -46,7 +47,11 @@ XEVY_BASE=https://your-xevy-host.example.com npx @xevy/skills -y
46
47
 
47
48
  Then:
48
49
 
49
- 1. **OAuth** — In Cursor: Settings → MCP → `xevy` → connect. In Claude Code: browser sign-in on first MCP use.
50
+ 1. **OAuth** — Authorize `xevy` in every installed client:
51
+ - **Codex:** `codex mcp login xevy`, then restart Codex or start a new session.
52
+ - **Claude Code:** run `/mcp`, select `xevy`, and complete authentication.
53
+ - **Cursor:** Settings → MCP → `xevy` → Connect.
54
+ - **Other clients:** use the client's MCP connect/OAuth action for `xevy`.
50
55
  2. **Setup** — Ask your agent to run the **Xevy setup** skill (or follow its steps) to link the repo and sync deploy config.
51
56
 
52
57
  ## Commands
@@ -67,7 +72,7 @@ Then:
67
72
  |------|-------------|
68
73
  | `-g, --global` | **(default)** User-global install (`~/.cursor/skills`, `~/.cursor/mcp.json`, …) |
69
74
  | `--project`, `--local`, `--no-global` | Install into the current project/repo instead of globally (opt-out) |
70
- | `-a, --agent` | Agent id (repeatable): `cursor`, `claude-code`, `codex`, … |
75
+ | `-a, --agent` | Agent id (repeatable). Explicit values override the default `cursor`, `claude-code`, `codex` targets. |
71
76
  | `-s, --skill` | Install specific skills only (passed to `npx skills`) |
72
77
  | `-y, --yes` | Skip prompts |
73
78
  | `--copy` | Copy skills instead of symlinking |
@@ -84,7 +89,7 @@ npx skills add ./node_modules/@xevy/skills -y
84
89
  ### Install MCP manually
85
90
 
86
91
  ```bash
87
- npx add-mcp https://xevy.ai/mcp -n xevy -a cursor -a claude-code -y
92
+ npx add-mcp https://xevy.ai/mcp -n xevy -a cursor -a claude-code -a codex -y
88
93
  ```
89
94
 
90
95
  ## Included skills
@@ -130,4 +135,4 @@ Version manifest for clients: `GET /api/agent-packages` on your Xevy host.
130
135
 
131
136
  - One-command setup: `npx @xevy/cli setup` (`@xevy/cli`)
132
137
  - MCP docs: https://xevy.ai/docs/mcp
133
- - Claude Code setup guide: https://xevy.ai/claude
138
+ - Coding-agent MCP setup guide: https://xevy.ai/docs/mcp
package/bin/cli.mjs CHANGED
@@ -26,7 +26,7 @@ const here = path.dirname(fileURLToPath(import.meta.url));
26
26
  const pkgRoot = path.resolve(here, "..");
27
27
 
28
28
  const DEFAULT_BASE = "https://xevy.ai";
29
- const DEFAULT_AGENTS = ["cursor", "claude-code"];
29
+ const DEFAULT_AGENTS = ["cursor", "claude-code", "codex"];
30
30
 
31
31
  function usage() {
32
32
  console.log(`Xevy skills + MCP installer
@@ -42,7 +42,8 @@ function usage() {
42
42
  Options:
43
43
  -g, --global (default) Install to user-global scope (~/.agents + global MCP)
44
44
  --project, --local Install to the current project/repo instead of global (opt-out)
45
- -a, --agent <name> Target agent (repeatable). Default: cursor, claude-code
45
+ -a, --agent <name> Target any agent supported by skills/add-mcp (repeatable)
46
+ Default: cursor, claude-code, codex
46
47
  -y, --yes Non-interactive
47
48
  --base <url> Xevy host (default: XEVY_BASE or ${DEFAULT_BASE})
48
49
  --manifest Prefer ${DEFAULT_BASE}/api/agent-packages for version check
@@ -69,6 +70,7 @@ function parseArgv(argv) {
69
70
  manifest: false,
70
71
  skillsOnly: false,
71
72
  agents: [],
73
+ agentsExplicit: false,
72
74
  skills: [],
73
75
  base: process.env.XEVY_BASE?.replace(/\/$/, "") || DEFAULT_BASE,
74
76
  rest: [],
@@ -132,6 +134,7 @@ function parseArgv(argv) {
132
134
  const v = args.shift();
133
135
  if (!v) die("Missing value for --agent");
134
136
  opts.agents.push(v);
137
+ opts.agentsExplicit = true;
135
138
  continue;
136
139
  }
137
140
  if (a === "-s" || a === "--skill") {
@@ -150,6 +153,7 @@ function parseArgv(argv) {
150
153
  }
151
154
 
152
155
  if (opts.agents.length === 0) opts.agents = [...DEFAULT_AGENTS];
156
+ opts.agents = [...new Set(opts.agents)];
153
157
  return opts;
154
158
  }
155
159
 
@@ -210,6 +214,32 @@ function recordInstallStamp(opts, version) {
210
214
  );
211
215
  }
212
216
 
217
+ function missingAgents(installedAgents, requestedAgents) {
218
+ const installed = new Set(Array.isArray(installedAgents) ? installedAgents : []);
219
+ return requestedAgents.filter((agent) => !installed.has(agent));
220
+ }
221
+
222
+ function printOAuthInstructions(opts) {
223
+ console.log(" • Authorize Xevy in each installed MCP client before using its tools:");
224
+ if (opts.agents.includes("codex")) {
225
+ console.log(" Codex: run `codex mcp login xevy`, then restart Codex or start a new session.");
226
+ }
227
+ if (opts.agents.includes("claude-code")) {
228
+ console.log(" Claude Code: run `/mcp`, select `xevy`, and complete authentication.");
229
+ }
230
+ if (opts.agents.includes("cursor")) {
231
+ console.log(" Cursor: open Settings → MCP → `xevy` and choose Connect.");
232
+ }
233
+ const otherAgents = opts.agents.filter(
234
+ (agent) => !DEFAULT_AGENTS.includes(agent),
235
+ );
236
+ if (otherAgents.length > 0) {
237
+ console.log(
238
+ ` ${otherAgents.join(", ")}: use the client's MCP connect/OAuth action for \`xevy\`.`,
239
+ );
240
+ }
241
+ }
242
+
213
243
  async function resolveLatestVersion(opts) {
214
244
  const bundled = readPackageVersion();
215
245
  let latest = bundled;
@@ -260,7 +290,12 @@ async function maybeReexecLatestCli(opts) {
260
290
  if (opts.manifest) forwardArgs.push("--manifest");
261
291
  if (opts.skillsOnly) forwardArgs.push("--skills-only");
262
292
  if (opts.base !== DEFAULT_BASE) forwardArgs.push("--base", opts.base);
263
- for (const a of opts.agents) forwardArgs.push("-a", a);
293
+ // Preserve a user's explicit target override, but let a newer package apply
294
+ // its own defaults. Forwarding this package's defaults would freeze old
295
+ // installations onto an outdated target list when a new default is added.
296
+ if (opts.agentsExplicit) {
297
+ for (const a of opts.agents) forwardArgs.push("-a", a);
298
+ }
264
299
  for (const s of opts.skills) forwardArgs.push("-s", s);
265
300
  forwardArgs.push(...opts.rest);
266
301
 
@@ -272,6 +307,7 @@ async function runCheck(opts) {
272
307
  const stamp = readInstalledStamp(process.cwd(), { global: opts.global });
273
308
  const { latest, source } = await resolveLatestVersion(opts);
274
309
  const installed = stamp?.version ?? null;
310
+ const missing = missingAgents(stamp?.agents, opts.agents);
275
311
  const cliVersion = readPackageVersion();
276
312
  const scope = opts.global ? "global" : "project";
277
313
 
@@ -280,6 +316,7 @@ async function runCheck(opts) {
280
316
  console.log(` Scope: ${scope}`);
281
317
  console.log(` Installed: ${installed ?? "(not recorded — run install or update)"}`);
282
318
  console.log(` Latest (${source}): ${latest}`);
319
+ console.log(` Target agents: ${opts.agents.join(", ")}`);
283
320
 
284
321
  if (!installed) {
285
322
  console.log(`\n⚠ No install stamp at ${stampPath(process.cwd(), { global: opts.global })}`);
@@ -287,8 +324,12 @@ async function runCheck(opts) {
287
324
  process.exit(1);
288
325
  }
289
326
 
290
- if (compareSemver(latest, installed) > 0) {
291
- console.log(`\n↻ Update available: ${installed} → ${latest}`);
327
+ const versionOutdated = compareSemver(latest, installed) > 0;
328
+ if (versionOutdated || missing.length > 0) {
329
+ if (versionOutdated) console.log(`\n↻ Update available: ${installed} → ${latest}`);
330
+ if (missing.length > 0) {
331
+ console.log(`${versionOutdated ? "" : "\n"}↻ Agent setup incomplete: ${missing.join(", ")}`);
332
+ }
292
333
  console.log(" Run: npx @xevy/skills update -y");
293
334
  process.exit(2);
294
335
  }
@@ -302,13 +343,20 @@ async function runUpdate(opts) {
302
343
  const stamp = readInstalledStamp(process.cwd(), { global: opts.global });
303
344
  const { latest } = await resolveLatestVersion(opts);
304
345
  const installed = stamp?.version ?? null;
346
+ const missing = missingAgents(stamp?.agents, opts.agents);
305
347
 
306
348
  if (installed && compareSemver(latest, installed) <= 0) {
307
- console.log(`✓ Xevy skills already at ${installed}`);
349
+ if (missing.length > 0) {
350
+ console.log(`→ Adding Xevy skills for: ${missing.join(", ")}…`);
351
+ installSkills(opts);
352
+ } else {
353
+ console.log(`✓ Xevy skills already at ${installed} for ${opts.agents.join(", ")}`);
354
+ }
308
355
  if (!opts.skillsOnly) {
309
356
  installMcp(opts);
310
357
  }
311
358
  recordInstallStamp(opts, installed);
359
+ if (!opts.skillsOnly && missing.length > 0) printOAuthInstructions({ ...opts, agents: missing });
312
360
  return;
313
361
  }
314
362
 
@@ -324,6 +372,7 @@ async function runUpdate(opts) {
324
372
  }
325
373
  recordInstallStamp(opts, latest);
326
374
  console.log(`\n✓ Xevy skills updated to ${latest}.`);
375
+ if (!opts.skillsOnly && missing.length > 0) printOAuthInstructions({ ...opts, agents: missing });
327
376
  }
328
377
 
329
378
  function forwardHelper(script, rest) {
@@ -360,18 +409,26 @@ async function main() {
360
409
  return;
361
410
  }
362
411
 
363
- if (opts.command === "skills" || opts.command === "install") {
412
+ if (opts.command === "skills") {
364
413
  installSkills(opts);
365
414
  recordInstallStamp(opts, readPackageVersion());
366
415
  }
367
416
 
368
- if (opts.command === "mcp" || opts.command === "install") {
417
+ if (opts.command === "mcp") {
369
418
  installMcp(opts);
370
419
  }
371
420
 
421
+ if (opts.command === "install") {
422
+ installSkills(opts);
423
+ installMcp(opts);
424
+ // Record coverage only after both installers succeed. A failed MCP write
425
+ // must not leave a stamp claiming that every requested agent is ready.
426
+ recordInstallStamp(opts, readPackageVersion());
427
+ }
428
+
372
429
  if (opts.command === "install") {
373
430
  console.log("\n✓ Xevy skills + MCP installed.");
374
- console.log(" • Complete OAuth in your agent (Cursor: Settings → MCP).");
431
+ printOAuthInstructions(opts);
375
432
  console.log(" • Run the setup skill in your repo (ask the agent to use the Xevy setup skill).");
376
433
  console.log(` • Docs: ${opts.base}/docs/mcp`);
377
434
  console.log(" • Later: npx @xevy/skills update -y");
@@ -379,6 +436,7 @@ async function main() {
379
436
  console.log("\n✓ Xevy skills installed.");
380
437
  } else if (opts.command === "mcp") {
381
438
  console.log("\n✓ Xevy MCP configured.");
439
+ printOAuthInstructions(opts);
382
440
  }
383
441
  }
384
442
 
package/package.json CHANGED
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "name": "@xevy/skills",
3
- "version": "0.1.21",
4
- "description": "Install Xevy agent skills and MCP for Cursor, Claude Code, and 50+ agents — one command via npx skills + add-mcp.",
3
+ "version": "0.1.22",
4
+ "description": "Install Xevy agent skills and MCP for Cursor, Claude Code, Codex, and other coding agents — one command via npx skills + add-mcp.",
5
5
  "type": "module",
6
6
  "keywords": [
7
7
  "xevy",
@@ -10,6 +10,7 @@
10
10
  "mcp",
11
11
  "cursor",
12
12
  "claude-code",
13
+ "codex",
13
14
  "sdlc",
14
15
  "deploy"
15
16
  ],
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: analyze-here
3
- description: Analyze a task locally in this Claude Code session — inspect the codebase, write the Analysis section + storyPoints via MCP, then self-review. Cloud equivalent is /xevy:analyze.
3
+ description: Analyze a task locally in the current coding-agent session — inspect the codebase, write the Analysis section + storyPoints via MCP, then self-review. Cloud equivalent is the analyze skill.
4
4
  argument-hint: "[task]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Bash mcp__xevy__find_task mcp__xevy__update_task mcp__xevy__list_specs mcp__xevy__list_subtasks mcp__xevy__create_task_note mcp__xevy__list_task_decisions mcp__xevy__create_decision mcp__xevy__answer_decision"
@@ -14,7 +14,7 @@ Run the analysis yourself in this session (no server agent). Mirror the Xevy ana
14
14
  Resolve from `$ARGUMENTS` or context (`mcp__xevy__find_task` / `find_task`). Read existing specs, subtasks, and notes.
15
15
 
16
16
  ## 1b. Decisions gate (BEFORE the work)
17
- Call `mcp__xevy__list_task_decisions`. If any decision is **open**, resolve them first — open decisions block AI actions. Present each open question to the user with **AskUserQuestion** (use its preset choices + explanations), record the answer with `mcp__xevy__answer_decision`, and fold the answers into the analysis. Never guess or bypass an open decision.
17
+ Call `mcp__xevy__list_task_decisions`. If any decision is **open**, resolve them first — open decisions block AI actions. Present each open question through the host's user-question mechanism (using its preset choices + explanations), record the answer with `mcp__xevy__answer_decision`, and fold the answers into the analysis. Never guess or bypass an open decision.
18
18
 
19
19
  ## 2. Inspect the codebase
20
20
  Read the relevant files (callers, callees, similar patterns). Identify whether the task involves UI changes; if so, study existing UI patterns and record a `UI/UX Flow` note (`create_task_note`). Base decisions on existing patterns — don't ask about every detail.
@@ -24,7 +24,7 @@ Use `mcp__xevy__update_task` to append a `## Analysis` section (do NOT replace t
24
24
  - **Overview:** plain-language summary, before/after behaviour, user impact, key decisions.
25
25
  - **Technical Implementation:** affected files/modules, API/data/UI changes, assumptions, risks, verification approach.
26
26
  - Estimate `storyPoints` (1 trivial · 2 small · 3 medium · 5 large · 8 XL · 13 epic) and set it.
27
- - **Open questions:** for a genuine ambiguity a human must resolve, don't bury it as an assumption — record it with `mcp__xevy__create_decision` (preset choices + explanation, optional follow-up per choice). It will gate the next step until answered. If the user is present, you may instead ask via **AskUserQuestion** and persist the result with `create_decision` + `answer_decision`.
27
+ - **Open questions:** for a genuine ambiguity a human must resolve, don't bury it as an assumption — record it with `mcp__xevy__create_decision` (preset choices + explanation, optional follow-up per choice). It will gate the next step until answered. If the user is present, use the host's user-question mechanism and persist the result with `create_decision` + `answer_decision`.
28
28
 
29
29
  ## 4. Review & correct (MANDATORY)
30
30
  Re-read your Analysis against the actual code: is every claim accurate, are affected files real, are risks/ACs concrete? Fix any gap or inaccuracy before finishing. Then call `mcp__xevy__list_task_decisions` once more (post-check): if any decision is still open, surface it to the user. Report a PASS/PARTIAL verdict, what you corrected, any open decisions, and the task URL. Do not push the user to plan/execute next.
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: create-xevy-task
3
- description: Capture the current Claude Code session's AI work into a reviewable Xevy task — document the user's goal + requirements and everything the AI changed (files, branch, PR, preview), create the task, attach PR/branch metadata, and move it to for_review or for_qa so a human developer or QA can pick it up. Also creates an ordinary new task when there is nothing to capture. Use /xevy:create-xevy-task.
3
+ description: Capture the current coding-agent session's AI work into a reviewable Xevy task — document the user's goal + requirements and everything the AI changed (files, branch, PR, preview), create the task, attach PR/branch metadata, and move it to for_review or for_qa so a human developer or QA can pick it up. Also creates an ordinary new task when there is nothing to capture.
4
4
  argument-hint: "[title or description] [--review|--qa]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Bash mcp__xevy__find_task mcp__xevy__list_tasks mcp__xevy__get_project_workbreakdown mcp__xevy__create_task mcp__xevy__update_task mcp__xevy__update_task_status mcp__xevy__create_task_note mcp__xevy__get_deploy_status mcp__xevy__list_task_decisions mcp__xevy__create_decision mcp__xevy__answer_decision"
@@ -44,7 +44,7 @@ Check `list_tasks`/`find_task` once more before creating to avoid a duplicate.
44
44
  ## 4b. Raise open questions (when ambiguous)
45
45
  After the task exists, if scope, approach, category/epic placement, or trade-offs are **genuinely ambiguous** — don't bury them in the description or guess:
46
46
  - Call `mcp__xevy__create_decision` with preset choices + explanations (optional follow-up per choice).
47
- - If the user is present, you may ask via **AskUserQuestion** first, then persist with `create_decision` + `answer_decision`.
47
+ - If the user is present, use the host's user-question mechanism first, then persist the answer with `create_decision` + `answer_decision`.
48
48
  - Open questions gate `/xevy:mark-ready`, analyze, plan, and execute until answered — keep them few and high-signal.
49
49
  - In **capture mode**, only raise decisions for real ambiguities in requirements or review routing — not for documented AI work.
50
50
  - In **ordinary mode**, this is the primary way to capture unresolved intake ambiguities.
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: execute-here
3
- description: Execute a task locally in this Claude Code session — worktree, implement against the plan, run tests, commit, push, open a PR, deploy a preview, then review/correct. Cloud equivalent is /xevy:execute.
3
+ description: Execute a task locally in the current coding-agent session — worktree, implement against the plan, run tests, commit, push, open a PR, deploy a preview, then review/correct. Cloud equivalent is the execute skill.
4
4
  argument-hint: "[task]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Edit Write Bash Agent mcp__xevy__find_task mcp__xevy__update_task mcp__xevy__update_task_status mcp__xevy__list_specs mcp__xevy__list_subtasks mcp__xevy__update_subtask mcp__xevy__list_sprints mcp__xevy__assign_task_to_sprint mcp__xevy__deploy_preview mcp__xevy__get_deploy_status mcp__xevy__list_task_decisions mcp__xevy__create_decision mcp__xevy__answer_decision"
@@ -12,7 +12,7 @@ You are the engine: implement the task in this session, following the Xevy workt
12
12
 
13
13
  ## 1. Resolve + gates
14
14
  - Resolve the task (`$ARGUMENTS` / context). Read its `## Analysis`, plan spec, and subtasks. If none and non-trivial, suggest `/xevy:plan-here` or ask whether to plan-and-execute or execute directly.
15
- - **Decisions gate (BEFORE implementing):** call `mcp__xevy__list_task_decisions`. If any decision is **open**, resolve it first — open decisions block execution. Present each via **AskUserQuestion**, record with `mcp__xevy__answer_decision`, and fold the answer into the work (if a relevant subtask is already done, the answer may warrant a new subtask). Never guess or bypass.
15
+ - **Decisions gate (BEFORE implementing):** call `mcp__xevy__list_task_decisions`. If any decision is **open**, resolve it first — open decisions block execution. Present each through the host's user-question mechanism, record with `mcp__xevy__answer_decision`, and fold the answer into the work (if a relevant subtask is already done, the answer may warrant a new subtask). Never guess or bypass.
16
16
  - Sprint gate: same as cloud — if an active sprint exists and the task isn't on it, require `## Analysis` + `storyPoints` > 0 then `assign_task_to_sprint`, else stop and ask.
17
17
 
18
18
  ## 2. Worktree protocol
@@ -22,13 +22,13 @@ You are the engine: implement the task in this session, following the Xevy workt
22
22
 
23
23
  ## 3. Implement
24
24
  - Move the task to `in_progress` (`update_task_status`) when implementation begins — mirrors the cloud `execute`, which flips to `in_progress` at start (and to `for_review` on success).
25
- - Work the subtasks in order with Read/Grep/Edit/Write (dispatch builder/frontend subagents via `Agent` with the worktree cwd when useful). Keep changes surgical, match existing style. Update subtask status via MCP as you go.
25
+ - Work the subtasks in order with the host's read/search/edit tools. When supported and useful, dispatch builder/frontend subagents with the worktree as their working directory. Keep changes surgical, match existing style, and update subtask status via MCP as you go.
26
26
  - **Frontend/UI changes:** before editing any UI (pages, components, app shell, forms, styling), follow the `/xevy:xevy-design` skill so the change matches the project design system (`./.xevy/design.md`).
27
- - If you hit a genuine ambiguity a human must resolve mid-implementation, raise it with `mcp__xevy__create_decision` (preset choices) rather than guessing — or, if the user is present, ask via **AskUserQuestion** and persist with `create_decision` + `answer_decision`. Open decisions block further AI steps until answered.
27
+ - If you hit a genuine ambiguity a human must resolve mid-implementation, raise it with `mcp__xevy__create_decision` (preset choices) rather than guessing — or, if the user is present, use the host's user-question mechanism and persist with `create_decision` + `answer_decision`. Open decisions block further AI steps until answered.
28
28
  - Validate after each step (type-check → lint → targeted tests).
29
29
 
30
30
  ## 4. Review & correct (MANDATORY)
31
- - Run the full project checks. Trace every requirement to code; check "must NOT" constraints. Dispatch the **reviewer** subagent for an independent PASS/PARTIAL/FAIL with file:line evidence. Fix everything it flags; re-run until green.
31
+ - Run the full project checks. Trace every requirement to code and check "must NOT" constraints. When the host supports subagents, dispatch a **reviewer** for an independent PASS/PARTIAL/FAIL with file:line evidence; otherwise perform a separate review pass yourself. Fix everything it flags and re-run until green.
32
32
 
33
33
  ## 5. Push, PR, deploy
34
34
  - Commit + push the task branch. Open/refresh the PR and ALWAYS give the user the PR/compare link. Do NOT merge automatically.
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: fix-here
3
- description: Apply a fix/rework to a task locally in this Claude Code session — reuse the task worktree, rework, run tests, push, refresh the PR, redeploy preview, then review. Cloud equivalent is /xevy:fix.
3
+ description: Apply a fix/rework to a task locally in the current coding-agent session — reuse the task worktree, rework, run tests, push, refresh the PR, redeploy preview, then review. Cloud equivalent is the fix skill.
4
4
  argument-hint: "[task]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Edit Write Bash Agent mcp__xevy__find_task mcp__xevy__update_task mcp__xevy__list_subtasks mcp__xevy__update_subtask mcp__xevy__deploy_preview mcp__xevy__get_deploy_status"
@@ -12,10 +12,10 @@ allowed-tools: "Read Grep Edit Write Bash Agent mcp__xevy__find_task mcp__xevy__
12
12
  Resolve the task (`$ARGUMENTS` / context). Find the EXISTING worktree/branch for it (`git worktree list`, match by task code). The worktree MUST already exist — do not create a new branch; if none, stop and ask. Merge the latest source branch into the feature branch first to avoid divergence.
13
13
 
14
14
  ## 2. Rework
15
- Make the requested fix in the worktree (Read/Grep/Edit/Write, or builder subagent with the worktree cwd). Keep it surgical; root-cause the issue rather than patching symptoms. Update subtasks via MCP.
15
+ Make the requested fix in the worktree with the host's read/search/edit tools, or use a builder subagent with the worktree as its working directory when the host supports one. Keep it surgical; root-cause the issue rather than patching symptoms. Update subtasks via MCP.
16
16
 
17
17
  ## 3. Review & correct (MANDATORY)
18
- Run the full project checks. Verify the fix resolves the issue with no regressions; check "must NOT" constraints. Dispatch the **reviewer** subagent (PASS/PARTIAL/FAIL + file:line). Fix anything flagged; re-run until green.
18
+ Run the full project checks. Verify the fix resolves the issue with no regressions and check "must NOT" constraints. When supported, dispatch a **reviewer** subagent (PASS/PARTIAL/FAIL + file:line); otherwise perform a separate review pass yourself. Fix anything flagged and re-run until green.
19
19
 
20
20
  **Frontend/UI changes:** before editing any UI (pages, components, app shell, forms, styling), follow the `/xevy:xevy-design` skill so the change matches the project design system (`./.xevy/design.md`).
21
21
 
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: merge-here
3
- description: Merge a task's branch locally in this Claude Code session — merge the feature branch into the current branch for review (no push), run tests, resolve conflicts, then review. Cloud equivalent is /xevy:merge.
3
+ description: Merge a task's branch locally in the current coding-agent session — merge the feature branch into the current branch for review (no push), run tests, resolve conflicts, then review. Cloud equivalent is the merge skill.
4
4
  argument-hint: "[task]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Edit Bash mcp__xevy__find_task mcp__xevy__deploy_main mcp__xevy__get_deploy_status"
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: move-to-qa-here
3
- description: Move a task to QA locally in this Claude Code session — write/update non-technical, UI-verifiable QA Notes, set status for_qa, then review. Cloud equivalent is /xevy:move-to-qa.
3
+ description: Move a task to QA locally in the current coding-agent session — write/update non-technical, UI-verifiable QA Notes, set status for_qa, then review. Cloud equivalent is the move-to-qa skill.
4
4
  argument-hint: "[task]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Bash mcp__xevy__find_task mcp__xevy__list_task_notes mcp__xevy__create_task_note mcp__xevy__update_task_note mcp__xevy__update_task_status"
@@ -1,6 +1,6 @@
1
1
  ---
2
2
  name: plan-here
3
- description: Plan a task locally in this Claude Code session — dispatch scout + plan-reviewer subagents, write the plan spec + ordered subtasks via MCP, then review. Cloud equivalent is /xevy:plan.
3
+ description: Plan a task locally in the current coding-agent session — dispatch scout + plan-reviewer subagents when supported, write the plan spec + ordered subtasks via MCP, then review. Cloud equivalent is the plan skill.
4
4
  argument-hint: "[task]"
5
5
  disable-model-invocation: true
6
6
  allowed-tools: "Read Grep Bash Agent mcp__xevy__find_task mcp__xevy__create_spec mcp__xevy__update_spec mcp__xevy__list_specs mcp__xevy__create_subtask mcp__xevy__list_subtasks mcp__xevy__list_task_decisions mcp__xevy__create_decision mcp__xevy__answer_decision mcp__xevy__get_process_owners mcp__xevy__update_task_status"
@@ -11,14 +11,14 @@ allowed-tools: "Read Grep Bash Agent mcp__xevy__find_task mcp__xevy__create_spec
11
11
  Run planning yourself in this session.
12
12
 
13
13
  ## 1. Resolve + recon
14
- Resolve the task (`find_task` (by code)); it should have a `## Analysis` section (else suggest `/xevy:analyze-here`). **Decisions gate (BEFORE the work):** call `mcp__xevy__list_task_decisions`; if any decision is **open**, resolve it first (open decisions block AI actions) — present each via **AskUserQuestion**, record with `mcp__xevy__answer_decision`, and reflect it in the plan. Dispatch the **scout** subagent (`Agent`) for codebase recon — relevant files, callers/callees, similar patterns, baseline health.
14
+ Resolve the task (`find_task` by code); it should have a `## Analysis` section (otherwise suggest the `analyze-here` skill). **Decisions gate (BEFORE the work):** call `mcp__xevy__list_task_decisions`; if any decision is **open**, resolve it first (open decisions block AI actions) — present each through the host's user-question mechanism, record with `mcp__xevy__answer_decision`, and reflect it in the plan. When the host supports subagents, dispatch a **scout** for codebase recon — relevant files, callers/callees, similar patterns, baseline health. Otherwise perform the same recon in the current session.
15
15
 
16
16
  ## 2. Write the plan
17
17
  - Create a plan spec (`mcp__xevy__create_spec`, kind `plan`) with step-by-step changes (file paths + line ranges) and a trace matrix (requirement → files → strategy → test).
18
18
  - Create ordered subtasks (`mcp__xevy__create_subtask`) grouping the work into reviewable pieces, each end-to-end (frontend + backend for one feature together).
19
19
 
20
20
  ## 3. Plan-review
21
- Dispatch the **plan-reviewer** subagent to cross-check every matrix item vs requirements (no orphans/gaps), verify cross-flow risks and "must NOT" constraints are covered, and that tests hit the exact branches. Fold its findings back into the spec/subtasks.
21
+ When the host supports subagents, dispatch a **plan-reviewer** to cross-check every matrix item vs requirements (no orphans/gaps), verify cross-flow risks and "must NOT" constraints are covered, and confirm tests hit the exact branches. Otherwise perform a separate self-review pass with the same checks. Fold the findings back into the spec/subtasks.
22
22
 
23
23
  When planning surfaces a genuine open question a human must resolve, record it with `mcp__xevy__create_decision` (preset choices + explanation) rather than guessing — it gates execution until answered.
24
24
 
@@ -8,26 +8,26 @@ allowed-tools: "Read Write Edit Bash mcp__xevy__get_current_user mcp__xevy__list
8
8
 
9
9
  # /xevy:xevy-setup — connect this repo to Xevy
10
10
 
11
- You are bootstrapping the Xevy integration for the current git repository. Work locally with the user's own tools; use the Xevy MCP (`mcp__xevy__*`) for all server state. Optional argument `$ARGUMENTS` may pre-fill `workspace/project`.
11
+ Bootstrap the Xevy integration for the current git repository. Work locally with the host coding agent's tools and use the Xevy MCP for all server state. Clients may display a tool as `xevy.get_current_user`, `mcp__xevy__get_current_user`, or a similar namespaced form; the short tool names below refer to that same Xevy MCP. An optional skill argument may pre-fill `workspace/project`.
12
12
 
13
13
  ## 0. Preconditions — MCP connection (OAuth)
14
14
  If MCP is not configured yet, run `npx @xevy/skills` from the repo root first.
15
15
  The Xevy MCP server authenticates via **OAuth** when configured (run `npx @xevy/skills` to add it) — no API key needed for MCP tools.
16
- - Call `mcp__xevy__get_current_user` to confirm the connection works.
17
- - If it returns "OAuth authorization required" (or similar), tell the user to complete Xevy OAuth (Cursor: Settings → MCP → xevy → connect; Claude Code: browser sign-in on first connect), then re-run. To target a non-default host, set `XEVY_BASE` (e.g. `https://xevy.vyte.dev`) and restart. Do not proceed until `get_current_user` succeeds.
16
+ - Call `get_current_user` to confirm the connection works.
17
+ - If it returns "OAuth authorization required" (or similar), stop and give the recovery step for the active client: Codex → run `codex mcp login xevy`, then restart Codex or start a new session; Claude Code → run `/mcp`, select `xevy`, and authenticate; Cursor → Settings → MCP → `xevy` → Connect; another client → use its MCP connect/OAuth action for `xevy`, then restart or reload that client if required. Re-run `get_current_user` after authentication. To target a non-default host, set `XEVY_BASE` (for example `https://xevy.vyte.dev`), reinstall the MCP configuration, and restart the client. Do not proceed until the call succeeds.
18
18
 
19
19
  ## 0.5. Agent skills (optional but recommended)
20
20
  - Run `npx @xevy/skills check` from the repo root. If an update is available, run `npx @xevy/skills update -y` (refreshes SDLC skills; MCP tools stay live on the server).
21
- - Skip if the user only uses Claude Code `/xevy:*` commands from the plugin and does not use Cursor/Codex skills.
21
+ - Skip only when the user explicitly does not want local skill updates.
22
22
 
23
23
  ## 1. Detect the repo
24
24
  - Run `git rev-parse --show-toplevel` for the repo root and `git remote get-url origin` for the remote. If there is no git repo or no `origin`, ask the user how to proceed.
25
25
  - If `.xevy/project.json` already exists, read it and treat this as a re-link: confirm with the user before overwriting.
26
26
 
27
27
  ## 2. Pick workspace + project
28
- - If `$ARGUMENTS` is `workspace/project`, resolve those slugs directly. Otherwise:
29
- - `mcp__xevy__list_workspaces` → if more than one, ask the user to choose.
30
- - `mcp__xevy__list_projects` for the chosen workspace. If none exist, offer to create one with `mcp__xevy__create_project` (ask for a name). If several exist, ask which.
28
+ - If the provided skill argument is `workspace/project`, resolve those slugs directly. Otherwise:
29
+ - `list_workspaces` → if more than one, ask the user to choose.
30
+ - `list_projects` for the chosen workspace. If none exist, offer to create one with `create_project` (ask for a name). If several exist, ask which.
31
31
  - Confirm the resolved `workspaceIdentifier` + `projectIdentifier` with the user before writing anything.
32
32
 
33
33
  ## 3. Link the repo + write local config
@@ -41,7 +41,7 @@ The Xevy MCP server authenticates via **OAuth** when configured (run `npx @xevy/
41
41
  Use a real ISO-8601 timestamp from `node -e "console.log(new Date().toISOString())"` (portable across macOS/Linux/Windows). Never put the API key in this file.
42
42
 
43
43
  ## 4. Deploy defaults — Dockerfile-first
44
- - Inspect the repo root: if a `Dockerfile` exists → `mcp__xevy__update_project_preview_settings({ buildPack: "dockerfile", dockerfilePath: "Dockerfile", ports: [<detected>] })`. Only if there is NO Dockerfile, use `docker_compose` (with `dockerComposeLocation`) when a compose file exists, else leave `nixpacks`. Detect ports from the code (3000 Node, 8000 Python, 8080 Go).
44
+ - Inspect the repo root: if a `Dockerfile` exists → `update_project_preview_settings({ buildPack: "dockerfile", dockerfilePath: "Dockerfile", ports: [<detected>] })`. Only if there is NO Dockerfile, use `docker_compose` (with `dockerComposeLocation`) when a compose file exists, else leave `nixpacks`. Detect ports from the code (3000 Node, 8000 Python, 8080 Go).
45
45
 
46
46
  ## 5. Sync environment variables (secrets never exposed)
47
47
  - Copy any **missing** env keys from the local `.env` into the project's deploy config by running the bundled helper — it reads the file in-process and prints only key names:
@@ -50,7 +50,7 @@ The Xevy MCP server authenticates via **OAuth** when configured (run `npx @xevy/
50
50
  - Auth note: the env-sync and repo-link helpers call the REST API (not MCP) and read an API key from `~/.config/xevy/credentials` (or `XEVY_API_KEY`). MCP itself uses OAuth; these two local helpers need a REST key. If none is present, tell the user to create one at Xevy → Settings → API keys, then re-run just these steps.
51
51
 
52
52
  ## 6. Bootstrap structure (optional)
53
- - If the project has no categories yet (`mcp__xevy__get_project_workbreakdown`), offer to create starter categories/epics — but do not create tasks. (Or tell the user to run a dedicated bootstrap later.)
53
+ - If the project has no categories yet (`get_project_workbreakdown`), offer to create starter categories/epics — but do not create tasks. (Or tell the user to run a dedicated bootstrap later.)
54
54
 
55
55
  ## 7. Report
56
- Summarise: workspace/project, repo link, deploy build pack + ports, count of env keys added (names only), and the next commands they can run (`/xevy:analyze`, `/xevy:execute`, `/xevy:deploy`). Do not print secrets.
56
+ Summarise: workspace/project, repo link, deploy build pack + ports, count of env keys added (names only), and the next Xevy skills they can run (`analyze`, `execute`, `deploy`; slash-command names may be available in some clients). Do not print secrets.