You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 4b57d11
Browse filesBrowse the repository at this point in the historyBrowse files
**Please do not open public issues for security vulnerabilities.**
13
+
14
+
Report vulnerabilities privately via GitHub Security Advisories:
15
+
16
+
👉 [Report a vulnerability](https://github.com/effectorHQ/.github/security/advisories/new)
17
+
18
+
Or email: **security@effectorhq.dev***(monitored, response within 72 hours)*
19
+
20
+
### What to include
21
+
22
+
- Package name and version affected
23
+
- Description of the vulnerability
24
+
- Steps to reproduce
25
+
- Potential impact
26
+
- Any suggested fix (optional)
27
+
28
+
### What to expect
29
+
30
+
1.**Acknowledgement** within 72 hours
31
+
2.**Assessment** within 7 days — we'll confirm scope and severity
32
+
3.**Fix and disclosure** — we coordinate a release and CVE if applicable, then publish a security advisory
33
+
34
+
We follow [Responsible Disclosure](https://en.wikipedia.org/wiki/Responsible_disclosure) and will credit researchers who report valid vulnerabilities (unless you prefer to remain anonymous).
35
+
36
+
## Scope
37
+
38
+
All packages under the `@effectorhq` npm scope and repositories in the [effectorHQ GitHub organization](https://github.com/effectorHQ) are in scope.
39
+
40
+
Note: effectorHQ packages are **static analysis tools** — they parse and validate files but do not execute agent tools or make network requests at runtime. The primary threat surface is path traversal and malicious input in parsed files.
npx @effectorhq/graph spectrum # visualize the 36-type capability map
37
39
```
38
40
39
41
Quick read: [Typed AI Agent Tools](https://github.com/effectorHQ/docs/blob/main/blog/typed-ai-agent-tools.md) · [v1.0 release announcement](https://github.com/effectorHQ/.github/discussions/8)
@@ -75,6 +77,7 @@ The core thesis: AI agent capabilities need types. Today you chain two skills an
75
77
| Project | What it does | Status |
76
78
|---------|-------------|--------|
77
79
|[`create-effector`](https://github.com/effectorHQ/create-effector)|`npx @effectorhq/create-effector` — scaffold any Effector type in seconds |[](https://npmjs.com/package/@effectorhq/create-effector)|
80
+
|[`studio`](https://github.com/effectorHQ/studio)|**Visual tool builder** — `npx @effectorhq/studio` opens a browser UI with live TOML/SKILL preview, 4-target compilation, and one-click scaffold |[](https://npmjs.com/package/@effectorhq/studio)|
78
81
|[`linear-skill`](https://github.com/effectorHQ/linear-skill)|**Reference implementation** — production-ready skill, Grade A eval, zero lint errors | v1.0.0 |
@@ -84,7 +87,7 @@ The core thesis: AI agent capabilities need types. Today you chain two skills an
84
87
| Project | What it does | Status |
85
88
|---------|-------------|--------|
86
89
|[`effector-compose`](https://github.com/effectorHQ/effector-compose)| Build agent pipelines, type-check them, emit to any runtime · `npm i @effectorhq/compose`|[](https://npmjs.com/package/@effectorhq/compose)|
87
-
|[`effector-graph`](https://github.com/effectorHQ/effector-graph)| Capability graph visualization and path-finding · `npm i @effectorhq/graph`|[](https://npmjs.com/package/@effectorhq/graph)|
0 commit comments