confluence-assistant
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to auto-detect workspace configuration by scanning various files, which creates an indirect prompt injection surface.
- Ingestion points: The skill reads from workspace documentation (e.g.,
.cursor/rules/confluence-config.mdc,AGENTS.md, project READMEs) and searches user-accessible Confluence pages as specified in the README.md file. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious instructions embedded within the documentation or Confluence pages it scans.
- Capability inventory: The agent can search, create, update, and comment on Confluence pages using Atlassian MCP tools.
- Sanitization: While the skill validates the format of Cloud IDs and URLs, it does not specify any sanitization or filtering of the natural language content it processes from external sources or workspace files.
- [NO_CODE]: The skill consists of documentation and configuration instructions only; it does not include any Python, Node.js, or shell scripts.
Audit Metadata