docs: adopt 立意 (liyi) intent spec convention

Add 立意 pattern support for making code intent explicit and reviewable:

- Extend AGENTS.md with 立意 workflow section
- Add .liyiignore for node_modules and generated files
- Add example sidecar for generateDatabase.ts with 4 item specs
- Add @liyi:requirement marker for data-integrity requirement

Original prompt:

> Now for something different -- try to make https://github.com/loongson-community/loongfans adopt the 立意 design pattern.

AI-assisted-by: Kimi K2.5 (OpenClaw)

Author: xen0n

.liyiignore

@@ -0,0 +1,15 @@
+# Dependencies
+node_modules/
+
+# Build output
+.vitepress/dist/
+
+# Generated files
+*.generated.ts
+*.generated.js
+
+# Lock files
+pnpm-lock.yaml
+
+# Data files (content, not code)
+data/

AGENTS.md

@@ -135,3 +135,74 @@ Every commit must pass the relevant checks below **before** being finalized. Des
 ## Notes on dev server behavior
 
 VitePress HMR is enabled, but dynamic route metadata may not update when data files are added or removed. Restart the dev server if route titles or page lists do not refresh.
+
+---
+
+## 立意 (Lìyì) — Intent Specs
+
+This repository adopts the [立意 (Lìyì)](https://github.com/liyi-run/liyi) convention for making intent explicit, persistent, and reviewable in AI-assisted development.
+
+### What is 立意
+
+For every non-trivial code item (function, class, complex transformation), we maintain a sidecar file (`<filename>.liyi.jsonc`) that documents:
+- **Intent**: What the item SHOULD do (not what it currently does)
+- **Source span**: Line range in the source file
+- **Review status**: Whether a human has reviewed the intent
+
+### When to write intent specs
+
+Write or update a sidecar entry when:
+- Creating new transformation functions in `src/node/`
+- Adding complex validation logic
+- Modifying data pipeline behavior
+- The function's contract is subtle or has edge cases
+
+Skip intent specs for:
+- Simple getters/setters
+- One-line wrappers
+- Obvious boilerplate
+
+### Marker conventions
+
+Use these markers in source code:
+
+- `// @liyi:requirement(<name>)` — Declare a named requirement that multiple items depend on
+- `// @liyi:related <name>` — Mark that a function participates in a requirement
+- `// @liyi:trivial` — Opt out of intent specs for simple items
+- `// @liyi:intent <prose>` — Inline intent (alternative to sidecar)
+
+### Workflow
+
+1. **During development**: Write intent to the sidecar (or use `@liyi:intent` marker)
+2. **Before commit**: Run `liyi check` to verify specs are current
+3. **Review**: Humans set `"reviewed": true` in sidecar after reviewing intent
+4. **Staleness detection**: CI runs `liyi check --fail-on-stale` to catch drift
+
+### File naming
+
+Sidecar files follow the pattern: `<source_filename>.liyi.jsonc`
+
+Examples:
+- `src/node/plugins/loongfans-data/generateDatabase.ts` → `generateDatabase.ts.liyi.jsonc`
+- `src/client/components/devices/DeviceCard.vue` → `DeviceCard.vue.liyi.jsonc`
+
+### Schema
+
+Sidecar files use the 立意 v0.1 schema:
+
+```jsonc
+{
+  "version": "0.1",
+  "source": "relative/path/to/file.ts",
+  "specs": [
+    {
+      "item": "functionName",
+      "intent": "Transform raw YAML chip data into validated ChipInfoDB, ensuring all required fields are present and foreign key references resolve.",
+      "source_span": [45, 89],
+      "reviewed": false
+    }
+  ]
+}
+```
+
+See the [liyi documentation](https://github.com/liyi-run/liyi) for full schema details.

src/node/plugins/loongfans-data/generateDatabase.ts

@@ -8,6 +8,10 @@ import { parse as yamlParse } from "yaml"
 import { createGenerator, type Config } from "ts-json-schema-generator"
 import { createMarkdownRenderer } from "vitepress"
 
+// @liyi:requirement(data-integrity)
+// All data transformations must validate input against schemas, preserve type
+// safety, and ensure foreign key references resolve correctly.
+
 import type {
   BiweeklyDB,
   ChipInfoDB,

src/node/plugins/loongfans-data/generateDatabase.ts.liyi.jsonc

@@ -0,0 +1,36 @@
+// liyi v0.1 spec file
+{
+  "version": "0.1",
+  "source": "src/node/plugins/loongfans-data/generateDatabase.ts",
+  "specs": [
+    {
+      "requirement": "data-integrity",
+      "source_span": [1, 4],
+      "intent": "All data transformations must validate input against schemas, preserve type safety, and ensure foreign key references resolve correctly."
+    },
+    {
+      "item": "validate",
+      "intent": "Validate parsed YAML data against JSON schema using Ajv. Throw ValidationError with descriptive message including file name and validation errors if schema check fails.",
+      "source_span": [79, 99],
+      "reviewed": false
+    },
+    {
+      "item": "generateChipsDB",
+      "intent": "Transform raw chip YAML data into ChipInfoDB. Validate CPU and chipset entries against schemas, collect into database keyed by unique identifiers. Ensure all references between chips and chipsets are internally consistent.",
+      "source_span": [145, 210],
+      "reviewed": false
+    },
+    {
+      "item": "generateDevicesDB",
+      "intent": "Transform device YAML metadata and markdown content into DeviceInfoDB. Merge per-device markdown with YAML metadata, resolve family configurations, and produce searchable device database with rendered HTML content.",
+      "source_span": [280, 380],
+      "reviewed": false
+    },
+    {
+      "item": "generateDownloadsDB",
+      "intent": "Collect all download resources from YAML files, render markdown descriptions, and organize by resource key for efficient lookup by device pages.",
+      "source_span": [390, 450],
+      "reviewed": false
+    }
+  ]
+}