team-coordination

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow instructs the agent to read _project_specs/team/state.md (and related team files) from the repository; those files are authored by team members/outsiders relative to the operating user, so their free-form markdown content can be ingested into the LLM context via the “read state.md / report active sessions / announcements” step.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)