@xevy/skills 0.1.19 → 0.1.21
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.
- package/.claude-plugin/plugin.json +1 -1
- package/package.json +1 -1
- package/skills/create-note/SKILL.md +27 -0
- package/skills/create-xevy-task/SKILL.md +1 -1
- package/skills/investigate/SKILL.md +4 -4
- package/skills/manage-notes/SKILL.md +32 -0
- package/skills/plan/SKILL.md +2 -1
- package/skills/plan-here/SKILL.md +2 -2
|
@@ -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.
|
|
5
|
+
"version": "0.1.21",
|
|
6
6
|
"author": {
|
|
7
7
|
"name": "Xevy"
|
|
8
8
|
},
|
package/package.json
CHANGED
|
@@ -0,0 +1,27 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: create-note
|
|
3
|
+
description: Create a personal note — a quick jotting, meeting note, snippet, or draft. Notes belong to you personally (not tied to any project or workspace) and are private by default; you can publish one to a stable public, no-login link. Use when the user asks to "jot down", "take a note", "make a note", "save this as a note", or "create a shareable note". For a project-scoped doc/RFC/runbook, prefer /xevy:create-document instead.
|
|
4
|
+
argument-hint: "[title or topic] [--publish]"
|
|
5
|
+
disable-model-invocation: true
|
|
6
|
+
allowed-tools: "Read mcp__xevy__get_current_user mcp__xevy__create_note mcp__xevy__publish_note"
|
|
7
|
+
---
|
|
8
|
+
|
|
9
|
+
# /xevy:create-note — create a personal note
|
|
10
|
+
|
|
11
|
+
Create a personal note and return a link. Notes are **per-user** — they belong to you and are NOT scoped to any project or workspace (unlike documents). They are `private` by default, or **published** to a stable no-login URL (`<XEVY_BASE>/n/<publicId>`). Optional argument `$ARGUMENTS` may carry the title/topic and a `--publish` flag.
|
|
12
|
+
|
|
13
|
+
## 1. Confirm the account
|
|
14
|
+
- Call `mcp__xevy__get_current_user` so the note is owned by the right account. No workspace or project is needed — notes are strictly personal and owner-only.
|
|
15
|
+
|
|
16
|
+
## 2. Decide title and body
|
|
17
|
+
- **Title:** from `$ARGUMENTS` or the user's request. Keep it concise and descriptive. Ask if it's unclear.
|
|
18
|
+
- **Body (markdown):** draft `contentMd` from the conversation/request. Don't fabricate facts — if the user only gave a title, create a short stub and say so. Use real markdown (headings, lists, code fences).
|
|
19
|
+
|
|
20
|
+
## 3. Create it
|
|
21
|
+
- Call `mcp__xevy__create_note` with `title` and `contentMd` (markdown) — no workspace/project parameters. Notes are created **private**.
|
|
22
|
+
- **Publish only on request:** if the user asked for a public/shareable link or passed `--publish`, call `mcp__xevy__publish_note` with the returned `noteId` afterwards. Confirm first — a published note is readable by anyone with the link, no login. The public id is stable (minted once), so the link never changes on later edits or re-publishes.
|
|
23
|
+
|
|
24
|
+
## 4. Report
|
|
25
|
+
- Give the user the note **title** and its in-app link from the tool's returned `appUrl` (shape: `<XEVY_BASE>/notes/<id>`, e.g. `https://xevy.ai/notes/<id>`) — this works even for a private note (it opens the note in their Notes app after login).
|
|
26
|
+
- If **published**, also share the no-login URL from the tool's returned `publicUrl` (shape: `<XEVY_BASE>/n/<publicId>`).
|
|
27
|
+
- Use the URLs the tool returns — do not construct them yourself. Never print secrets or paste private content into third-party tools.
|
|
@@ -59,7 +59,7 @@ After the task exists, if scope, approach, category/epic placement, or trade-off
|
|
|
59
59
|
No AI work to capture (or a fresh, unrelated task): create the task from `$ARGUMENTS`/the request. Capture the user's story, problem, expectation, and desired outcome in plain language — do NOT fabricate technical detail or a PR. Raise any genuine ambiguities with `create_decision` (section 4b). Leave status at `draft` — use `/xevy:mark-ready` to promote once open questions are answered. Do NOT set `ready` while any decision is still open. Skip the review/QA routing.
|
|
60
60
|
|
|
61
61
|
## 7. Report
|
|
62
|
-
Give the task **code + the `url` returned by the create/update tool** (always show this link so the user can open and review it), its **status**, any **open decisions** (call `list_task_decisions` if you raised questions), and a one-line summary of what was captured (PR link, branch, preview URL). Tell the user a developer/QA can now pick it up directly from Xevy (capture mode) or that they should answer open questions / run `/xevy:mark-ready
|
|
62
|
+
Give the task **code + the `url` returned by the create/update tool** (always show this link so the user can open and review it), its **status**, any **open decisions** (call `list_task_decisions` if you raised questions), and a one-line summary of what was captured (PR link, branch, preview URL). Tell the user a developer/QA can now pick it up directly from Xevy (capture mode) or that they should answer open questions and analyze/plan while draft (ordinary mode). After planning, AI Agent-owned Process Owner stages advance automatically; human-owned stages wait for the user to run `/xevy:mark-ready`.
|
|
63
63
|
|
|
64
64
|
## Rules
|
|
65
65
|
- Evidence before claims — gather PR/branch/changes/tests from git and the session; never invent them.
|
|
@@ -54,9 +54,9 @@ For non-SQL service connectors, run a **read** operation through the gateway ins
|
|
|
54
54
|
3. **Check params + cost:** `connector_describe` with the `serviceName` + `operation`.
|
|
55
55
|
4. **Run it:** `connector_execute` with `serviceName`, `operation`, and `params`. Large results come back as a summary + `resultRef` (fetch with `connector_fetch_result`); use `limit` to keep responses small.
|
|
56
56
|
|
|
57
|
-
**Shopify data connector —
|
|
57
|
+
**Shopify data connector — one flexible GraphQL surface (+ a liveness shortcut):** the connector exposes `graphql_query` (read-only) and `graphql_mutation` (guarded write), which cover the **entire** Shopify Admin GraphQL API — products, orders, inventory, customers, metafields, themes, redirects, everything. There are no typed per-resource ops to hunt for; you write the GraphQL document. The one exception is `shop_info`, a zero-param convenience read for the quickest "is the connection alive?" check. These prove store state without any local script or token handling. Example: to verify the Shopify connection works, `connector_execute serviceName:"shopify" operation:"shop_info"` and report the returned store name + domain.
|
|
58
58
|
|
|
59
|
-
> **Investigation stays read-only.** Shopify **write**
|
|
59
|
+
> **Investigation stays read-only.** `graphql_query` refuses mutation documents outright; a Shopify **write** goes through `graphql_mutation`, which is approval-guarded — a plain `connector_execute` returns a **dry-run preview** and never mutates. During investigation, never pass `approve:true`; a fix (if any) is a separate, user-approved step.
|
|
60
60
|
|
|
61
61
|
**Only fall back to an ephemeral script + env file** (below) when there is no connector for the target environment/service, or the check isn't a DB query / connector operation (SMTP handshake, internal HTTP route, config-flag presence, etc.).
|
|
62
62
|
|
|
@@ -190,7 +190,7 @@ After reporting:
|
|
|
190
190
|
|
|
191
191
|
### Store / e-commerce data (Shopify)
|
|
192
192
|
|
|
193
|
-
- **Preferred — Shopify connector:** if the project has an active `shopify` connector, investigate through it. `
|
|
193
|
+
- **Preferred — Shopify connector:** if the project has an active `shopify` connector, investigate through it. The connector exposes one flexible pair — `graphql_query` (read) and `graphql_mutation` (guarded write) — covering the whole Admin GraphQL API, plus `shop_info` as a zero-param liveness check (store name/domain/plan/currency). Use `shop_info` to confirm the connection, then write any Admin GraphQL document for catalogue, sales, or stock questions. Confirm params with `connector_describe`, run with `connector_execute` (`limit` to cap rows). No token handling, no script.
|
|
194
194
|
- **Read-only:** do not `approve` write ops while investigating — a plain call is a safe dry-run preview.
|
|
195
195
|
- **Fallback — no connector:** only then hit the Admin API with a token from the chosen env file via an ephemeral script; report status + redacted shape, never the token.
|
|
196
196
|
|
|
@@ -210,7 +210,7 @@ After reporting:
|
|
|
210
210
|
3. **No blind fixes** — investigate → report → wait for approval to fix.
|
|
211
211
|
4. **No secrets in chat** — not in scripts, logs, commits, or responses (connectors keep credentials in the gateway; never print row data that may contain PII).
|
|
212
212
|
5. **No committing** investigation scripts or env files.
|
|
213
|
-
6. **Prod caution / read-only investigation** — confirm with the user before `prod.env` or any write/mutate operation. SQL connectors are read-only-enforced (`connector_query` cannot mutate); service-connector writes (e.g. Shopify `
|
|
213
|
+
6. **Prod caution / read-only investigation** — confirm with the user before `prod.env` or any write/mutate operation. SQL connectors are read-only-enforced (`connector_query` cannot mutate); service-connector writes (e.g. Shopify `graphql_mutation`) are approval-guarded — during investigation never pass `approve:true`, keep to read ops and dry-run previews.
|
|
214
214
|
7. **Delete scripts** after every investigation run.
|
|
215
215
|
|
|
216
216
|
## Examples
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: manage-notes
|
|
3
|
+
description: Manage your personal notes end to end — list, find, read, edit, publish/unpublish, and delete. Use when the user wants to review their notes, update a note's body/title, publish a note to a public link or unpublish it, or remove a note. Notes belong to you personally and are not tied to any project or workspace. For creating a brand-new note, prefer /xevy:create-note.
|
|
4
|
+
argument-hint: "[what to do, e.g. 'list my notes', 'publish the standup note', 'delete the scratch note']"
|
|
5
|
+
disable-model-invocation: true
|
|
6
|
+
allowed-tools: "Read mcp__xevy__get_current_user mcp__xevy__list_notes mcp__xevy__get_note mcp__xevy__create_note mcp__xevy__update_note mcp__xevy__delete_note mcp__xevy__publish_note mcp__xevy__unpublish_note"
|
|
7
|
+
---
|
|
8
|
+
|
|
9
|
+
# /xevy:manage-notes — manage your personal notes
|
|
10
|
+
|
|
11
|
+
Full lifecycle management of your notes. Notes are **per-user** — they belong to you and are NOT scoped to any project or workspace (unlike documents): `private` by default, or **published** to a stable no-login URL (`<XEVY_BASE>/n/<publicId>`). You can only ever see and act on your own notes. `$ARGUMENTS` describes what to do.
|
|
12
|
+
|
|
13
|
+
## 1. Confirm the account
|
|
14
|
+
- `mcp__xevy__get_current_user` confirms whose notes you're working with. No workspace/project resolution is needed — notes are strictly personal and owner-only.
|
|
15
|
+
|
|
16
|
+
## 2. Find the note(s)
|
|
17
|
+
- **List:** `mcp__xevy__list_notes` → your own notes, newest-updated first. Pass `search` for a case-insensitive filter over title + body.
|
|
18
|
+
- **Read one:** `mcp__xevy__get_note` with the `noteId` for the full markdown `contentMd`.
|
|
19
|
+
- Match the user's description (title/topic) to a note; if several match, list the candidates and ask which one. Never guess at a destructive action.
|
|
20
|
+
|
|
21
|
+
## 3. Apply the change
|
|
22
|
+
- **Edit body / title** — `mcp__xevy__update_note`: set `contentMd` (markdown) and/or `title`. When revising, read the current `contentMd` first and preserve content the user didn't ask to change. Pass a `metadata` object to merge arbitrary keys.
|
|
23
|
+
- **Publish** — `mcp__xevy__publish_note`: makes the note readable by anyone with the link, no login. **Confirm first.** The public id is minted once and never changes, so re-publishing keeps the SAME link.
|
|
24
|
+
- **Unpublish** — `mcp__xevy__unpublish_note`: the public link stops resolving. The id is retained, so a later re-publish reuses the same URL.
|
|
25
|
+
|
|
26
|
+
## 4. Delete
|
|
27
|
+
- `mcp__xevy__delete_note` (soft delete). This is destructive — confirm the exact title with the user first. You can only delete your own notes.
|
|
28
|
+
|
|
29
|
+
## 5. Report
|
|
30
|
+
- State what changed (title, published/unpublished). Give the in-app link from the tool's returned `appUrl` (shape: `<XEVY_BASE>/notes/<id>`) — works even for a private note. When a note is **published**, also give the no-login link from the tool's returned `publicUrl` (shape: `<XEVY_BASE>/n/<publicId>`). Use the URLs the tool returns — never construct or guess them.
|
|
31
|
+
- For lists, show title · updated date · (public) — concise, scannable.
|
|
32
|
+
- Never paste private note content into third-party tools, and never print secrets.
|
package/skills/plan/SKILL.md
CHANGED
|
@@ -21,4 +21,5 @@ Resolve from `$ARGUMENTS` or context. The task should be analyzed first (a `## A
|
|
|
21
21
|
- Build a quick trace matrix: every requirement from the task/Analysis → a subtask/step that covers it. Flag orphans (requirements with no step) and gaps (steps with no requirement). Inspect the code with Read/Grep to confirm the plan is feasible and references real files.
|
|
22
22
|
- Check cross-flow risks (guards, context rehydration, effects, async side effects) are addressed, and that any "must NOT" constraints have explicit verification.
|
|
23
23
|
- **Fix** gaps: edit the spec / add or reorder subtasks via MCP. Re-review until the matrix has no orphans.
|
|
24
|
-
-
|
|
24
|
+
- Do **not** change the task status or start execution yourself. The server applies Process Owner automation when the run succeeds: AI Agent-owned stages advance automatically; human-owned stages wait for the user.
|
|
25
|
+
- Report a PASS/PARTIAL verdict, the gaps you closed, the resulting task status, and the task URL.
|
|
@@ -3,7 +3,7 @@ name: plan-here
|
|
|
3
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.
|
|
4
4
|
argument-hint: "[task]"
|
|
5
5
|
disable-model-invocation: true
|
|
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"
|
|
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"
|
|
7
7
|
---
|
|
8
8
|
|
|
9
9
|
# /xevy:plan-here — plan locally (then review)
|
|
@@ -23,4 +23,4 @@ Dispatch the **plan-reviewer** subagent to cross-check every matrix item vs requ
|
|
|
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
|
|
|
25
25
|
## 4. Review & correct (MANDATORY)
|
|
26
|
-
Confirm the final plan is complete and feasible against the actual code; fix any remaining gap before finishing. Post-check: call `mcp__xevy__list_task_decisions` — if any decision is still open, surface it. Report a PASS/PARTIAL verdict, the matrix summary, any open decisions, and the task URL.
|
|
26
|
+
Confirm the final plan is complete and feasible against the actual code; fix any remaining gap before finishing. Post-check: call `mcp__xevy__list_task_decisions` — if any decision is still open, surface it. Then call `mcp__xevy__get_process_owners`: if the task is still `draft` and the Refine owner is `ai`, automatically call `mcp__xevy__update_task_status({ taskId, status: "ready" })`; if Refine is human-owned, leave it draft for the user to move manually. Do not start execution as part of local planning. Report a PASS/PARTIAL verdict, the matrix summary, any open decisions, the resulting status, and the task URL.
|