Skip to main content

security

Implements secure coding practices for agent-native applications, focusing on input validation, SQL injection prevention, and data protection.

Install this skill

or
5/100

Security score

The security skill was audited on Jun 6, 2026 and we found 13 security issues across 3 threat categories, including 1 critical. Review the findings below before installing.

Categories Tested

Security Issues

critical line 71

Eval function call - arbitrary code execution

SourceSKILL.md
71- Never use `dangerouslySetInnerHTML`, `innerHTML`, `eval()`, or `document.write()` with user-controlled content.
high line 26

Template literal with variable interpolation in command context

SourceSKILL.md
26`<OPENAI_API_KEY>`, `${keys.SLACK_WEBHOOK}`, `sk-test-example`, or
medium line 65

Template literal with variable interpolation in command context

SourceSKILL.md
65await client.execute(`SELECT * FROM users WHERE id = '${id}'`);
medium line 24

Webhook reference - potential data exfiltration

SourceSKILL.md
24tokens, webhook URLs, signing secrets, private Builder/internal data, or customer
medium line 26

Webhook reference - potential data exfiltration

SourceSKILL.md
26`<OPENAI_API_KEY>`, `${keys.SLACK_WEBHOOK}`, `sk-test-example`, or
medium line 93

Webhook reference - potential data exfiltration

SourceSKILL.md
93- Never commit real keys, tokens, webhook URLs, signing secrets, or private
medium line 119

Webhook reference - potential data exfiltration

SourceSKILL.md
119- `scripts/guard-no-env-mutation.mjs` — bans `process.env.<KEY> = …` (and bracket / compound forms) anywhere in production code. On serverless, every warm container handles many concurrent requests in
medium line 230

Webhook reference - potential data exfiltration

SourceSKILL.md
230- [ ] No hardcoded API keys, tokens, webhook URLs, signing secrets, real
medium line 97

Access to .env file

SourceSKILL.md
97## User Credentials Are Per-User Data — Never `process.env`
medium line 108

Access to .env file

SourceSKILL.md
108On 2026-04-29 the previous one-arg `resolveCredential(key)` form fell back to `process.env[key]` and an unscoped global `settings` row, so every signed-in user inherited the deployment's credentials.
medium line 110

Access to .env file

SourceSKILL.md
110- `scripts/guard-no-env-credentials.mjs` — bans `process.env.<KEY>` reads in `packages/core/src/credentials/`, `secrets/`, `vault/`, and `templates/*/server/{lib,routes/api}/credential*` paths, except
medium line 119

Access to .env file

SourceSKILL.md
119- `scripts/guard-no-env-mutation.mjs` — bans `process.env.<KEY> = …` (and bracket / compound forms) anywhere in production code. On serverless, every warm container handles many concurrent requests in
medium line 232

Access to .env file

SourceSKILL.md
232- [ ] New env vars in `.env` only, not committed
Scanned on Jun 6, 2026
View Security Dashboard
Installation guide →