CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

minimal-web-baas-demo

Fast path for a minimal CloudBase Web + database demo (最小前后端 / 最小可用 fullstack / Lovable-like BaaS). Defaults to @cloudbase/js-sdk client CRUD (NoSQL app.database / PG app.rdb), MCP-only schema, preview-first, and forbids cloud functions unless secrets, cron/background jobs, or logic that security rules/RLS cannot express. Use for 搭一套 demo、留言板、Todo、Notes、Kanban, or when users say 带云函数+云数据库 but only need CRUD. NOT for production multi-service backends, CloudRun, WeChat Mini Programs, or tasks that truly need server secrets.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

The canonical home for this skill is minimal-web-baas-demo in TencentCloudBase/CloudBase-AI-Toolkit

SKILL.md
Quality
Evals
Security

Quality

Content

61%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A dense, high-signal fast-path skill: the hard rules, sniff order, and playbook are concrete and clearly sequenced, with named MCP tools, SDK calls, and an executable auth snippet. Its weaknesses are redundancy (the zero-cloud-function rule and sniff order are repeated multiple times, plus off-task partner-host meta content) and absent validation checkpoints in a database workflow — no verify step after schema creation or preview smoke-test before reporting the URL.

Suggestions

State the cloud-function prohibition once (Hard rule 2) and remove its re-statements in the sniff order, playbook step 6, and the paste block; move the 'WorkBuddy / partner packaging notes' and 'One-screen partner paste' sections to a separate reference file to cut roughly 25% of body tokens.

Add explicit validation checkpoints to the playbook: after MCP schema creation, verify the collection/table and permissions with a test read before wiring the frontend, and smoke-test list + add on the preview URL before reporting it (validate → fix → retry), which would also lift workflow_clarity above the database-operations cap.

Include one complete minimal CRUD example (list + add on a collection via app.database(), including the npm install @cloudbase/js-sdk step) so the core deliverable is copy-paste ready rather than API signatures.

DimensionReasoningScore

Conciseness

The core rules and playbook are tight (no library tutorials, no concepts Claude already knows), but the body is noticeably redundant and could be tightened: the cloud-function prohibition is restated four times (Hard rule 2, sniff-order step 7 "Cloud functions ← skip (count = 0)", playbook step 6 "Cloud function count for this path = 0", and the paste block "Do not create cloud functions for CRUD"), the sniff order appears twice (ordered list plus "One-screen partner paste"), and the ~25-line "WorkBuddy / partner packaging notes" section is host-meta content off the executing agent's critical path. That is more than the 'minor instances' of the score-4 anchor, fitting 'mostly efficient but includes some unnecessary explanation or could be tightened'.

3 / 5

Actionability

Concrete and mostly executable: named MCP tools ("writeNoSqlDatabaseStructure", "queryPgDatabase / managePgDatabase"), named SDK calls ("app.database() → db.collection(...).get()/add()/update()/watch(...)"), a copy-paste-ready auth snippet ("const { error } = await auth.signInAnonymously(); if (error) throw error"), and concrete playbook commands ("downloadTemplate", "npm install", "queryEnv(action=\"info\")"). Not score 5 because the core deliverable — wiring list + add — is given only as API signatures with no complete end-to-end example, and the js-sdk install command is implied rather than stated; minor gaps rather than missing key details, so above the pseudocode anchor.

4 / 5

Workflow Clarity

The sequence is the strongest part — a fixed "Capability sniff order" 0–7 marked "Do not reorder", a 6-step "Standard playbook", and explicit decision conditions for cloud functions ("(a) the logic cannot be expressed as database security rules / RLS, and (b) it needs server-side secrets"). However, this is a database-operations workflow and the rubric's feedback-loop guidance applies: there is no validate-then-proceed checkpoint after MCP schema creation (no verify that the collection/table and permissions are live before wiring the frontend) and no smoke-test of the preview before reporting the URL — only the 401 rule and the snippet's `if (error) throw error` offer partial error handling. Per the scoring note, missing validate→fix→retry loops in database-operation workflows caps this at 3 ('checkpoints missing or implicit'), despite the clear sequence.

3 / 5

Progressive Disclosure

No bundle files exist (no references/, scripts/, or assets/ directories), and the body keeps itself compact by deferring detail through well-signaled, one-level-deep sibling references: the "On-demand skills (fetch only when needed)" table maps each need to a specific path (e.g., "../cloudbase-document-database-web-sdk/SKILL.md"), reinforced by "Do not dump every CloudBase skills at session start" and the searchKnowledgeBase fallback. Not score 5 because the inline "WorkBuddy / partner packaging notes" and "One-screen partner paste" sections are in-file content that belongs behind a reference or trimmed, a minor organization gap matching 'most content is appropriately placed... minor organization gaps'.

4 / 5

Total

14

/

20

Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

An exemplar description by this rubric: it states concrete capabilities with named SDKs and tools, gives explicit bilingual 'use for' triggers including verbatim user phrasing, and closes with a 'NOT for' clause that fences off adjacent skills. Voice is impersonal (no first/second person), and although long, every clause carries information rather than padding.

DimensionReasoningScore

Specificity

The description lists multiple concrete, named capabilities: "@cloudbase/js-sdk client CRUD (NoSQL app.database / PG app.rdb), MCP-only schema, preview-first, and forbids cloud functions unless secrets, cron/background jobs, or logic that security rules/RLS cannot express" — specific APIs and tools rather than vague abstractions, matching the comprehensive-coverage anchor. It is dense but every clause is informational (no padding), so the verbosity guideline is not tripped; not score 4 because coverage of the skill's actions (data plane, schema plane, cloud-function gating, preview approach) is complete, not 'minor gaps'.

5 / 5

Completeness

Both halves are explicit: the 'what' opens the description ("Fast path for a minimal CloudBase Web + database demo... Defaults to @cloudbase/js-sdk client CRUD...") and the 'when' is a concrete trigger clause ("Use for 搭一套 demo、留言板、Todo、Notes、Kanban, or when users say 带云函数+云数据库 but only need CRUD") plus negative scope ("NOT for production multi-service backends, CloudRun, WeChat Mini Programs..."). This matches the anchor-5 good example pattern exactly; not score 4 because the 'when' is already fully explicit with concrete trigger phrases.

5 / 5

Trigger Term Quality

Natural user phrasing is comprehensively covered with bilingual synonyms and concrete app names: "搭一套 demo、留言板、Todo、Notes、Kanban", "最小前后端 / 最小可用 fullstack / Lovable-like BaaS", plus the verbatim user phrase "带云函数+云数据库 but only need CRUD". A user needing this skill would literally say these terms; not score 4 because synonym coverage (Chinese/English, app-type names) leaves no notable natural variants missing for the target audience.

5 / 5

Distinctiveness Conflict Risk

Clear niche (minimal CloudBase Web BaaS demo) with explicit disambiguation against the nearest competing skills: "NOT for production multi-service backends, CloudRun, WeChat Mini Programs, or tasks that truly need server secrets". Trigger terms (留言板/Todo/Kanban/Lovable-like) are distinctive to this niche; not score 4 because overlap risk with sibling CloudBase skills is explicitly fenced off rather than merely minor.

5 / 5

Total

20

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

TencentCloudBase/CloudBase-AI-Toolkit

Path

config/.claude/skills/minimal-web-baas-demo/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.