insforge-cli

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes examples and commands that embed secret values verbatim (e.g., env set with "ik_xxx", inline --env JSON, secrets get which returns decrypted values, and schedule headers using ${{secrets...}}), so an agent following it may be required to output or handle secret strings directly, creating exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill's create workflow runs "npx skills add insforge/agent-skills" during project creation, which fetches and executes remote package code that installs agent skills (i.e., external code that can directly control agent prompts/behavior) — this is a runtime external dependency that meets the criteria.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly exposes Stripe payment management: commands to set Stripe test/live secret keys, sync products/prices/customers/subscriptions, configure webhooks, and create/update/delete products and prices, plus admin customer and subscription operations. Managing secret keys and Stripe catalog/products/prices/webhooks is a payment-gateway integration (Stripe) — a specific financial execution capability per the rule.