clawsec-feed

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md explicitly fetches and parses a public, community-driven advisory feed (DEFAULT_FEED_URL = https://raw.githubusercontent.com/prompt-security/ClawSec/main/advisories/feed.json) and GitHub release artifacts, and it uses those untrusted third-party advisories to cross-reference installed skills and drive notifications/remediation actions, so third-party content can materially influence agent decisions.

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

• Potentially malicious external URL detected (high risk: 1.00). The install and runtime scripts fetch and install remote skill artifacts and the advisory feed from GitHub (e.g. https://api.github.com/repos/prompt-security/ClawSec/releases and https://raw.githubusercontent.com/prompt-security/ClawSec/main/advisories/feed.json), which are retrieved at runtime and can inject remote JSON/skill files that directly control agent messages/behavior or supply code to be installed.