eslint-plugin-ai-guard 1.1.9 โ 1.1.11
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/README.md +27 -2
- package/dist/cli/index.js +548 -27
- package/dist/cli/index.js.map +1 -1
- package/dist/index.js +481 -99
- package/dist/index.js.map +1 -1
- package/dist/index.mjs +481 -99
- package/dist/index.mjs.map +1 -1
- package/package.json +5 -4
package/README.md
CHANGED
|
@@ -28,11 +28,35 @@ npx ai-guard run # recommended preset (lowest noise)
|
|
|
28
28
|
npx ai-guard run --strict
|
|
29
29
|
npx ai-guard run --security
|
|
30
30
|
npx ai-guard init # auto-creates ESLint config for you
|
|
31
|
+
npx ai-guard init --dry-run
|
|
31
32
|
npx ai-guard doctor # diagnoses setup issues
|
|
32
33
|
npx ai-guard baseline # track only *new* issues going forward
|
|
33
34
|
```
|
|
34
35
|
That's it. **Zero configuration required.**
|
|
35
36
|
|
|
37
|
+
## ๐ค Set Up AI Agent Rules
|
|
38
|
+
|
|
39
|
+
Generate instruction files so Claude Code, Cursor, and GitHub Copilot
|
|
40
|
+
automatically avoid the 17 most common AI-generated anti-patterns:
|
|
41
|
+
|
|
42
|
+
```bash
|
|
43
|
+
npx ai-guard init-context
|
|
44
|
+
```
|
|
45
|
+
|
|
46
|
+
Follow the prompts to select your agent(s). Or generate all at once:
|
|
47
|
+
|
|
48
|
+
```bash
|
|
49
|
+
npx ai-guard init-context --all
|
|
50
|
+
```
|
|
51
|
+
|
|
52
|
+
This writes:
|
|
53
|
+
- `CLAUDE.md` โ read automatically by Claude Code
|
|
54
|
+
- `.cursorrules` โ read automatically by Cursor
|
|
55
|
+
- `.github/copilot-instructions.md` โ read automatically by GitHub Copilot
|
|
56
|
+
|
|
57
|
+
Your AI tools will now avoid these patterns before you even run the linter.
|
|
58
|
+
Use `--force` to regenerate after upgrading to a new version with new rules.
|
|
59
|
+
|
|
36
60
|
## ๐งช Real-World Usage Philosophy
|
|
37
61
|
|
|
38
62
|
`ai-guard` is designed for production adoption in existing codebases:
|
|
@@ -62,7 +86,8 @@ const users = await Promise.all(userIds.map(async (id) => {
|
|
|
62
86
|
|
|
63
87
|
### Terminal Output
|
|
64
88
|
|
|
65
|
-
|
|
90
|
+
|
|
66
91
|
|
|
67
92
|
*The terminal output above shows `ai-guard` catching multiple AI-generated anti-patterns in a single run.*
|
|
68
93
|
|
|
@@ -102,7 +127,7 @@ const users = await Promise.all(userIds.map(async (id) => {
|
|
|
102
127
|
Disallow dynamic `eval()` or `new Function()`.
|
|
103
128
|
- **`ai-guard/no-sql-string-concat`** (Warn in `recommended`, Error in `security`/`strict`)
|
|
104
129
|
Disallow variable concatenation/interpolation in SQL queries. AI tools frequently generate dangerous code enabling SQL injection.
|
|
105
|
-
- **`ai-guard/no-unsafe-deserialize`** (Warn)
|
|
130
|
+
- **`ai-guard/no-unsafe-deserialize`** (Warn in `recommended`/`security`, Error in `strict`)
|
|
106
131
|
Disallow `JSON.parse()` on likely untrusted inputs (like `req.body`) without visible validation.
|
|
107
132
|
- **`ai-guard/require-auth-middleware`** (Warn)
|
|
108
133
|
Enforce authentication middleware on Express/Fastify routes. AI tools frequently generate unprotected endpoints exposing sensitive data.
|