pm-organization-prd-reorganization
Pass
Audited by Gen Agent Trust Hub on Mar 24, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes external data sources (
docs/design/gdd.mdand./.claude/session/retrospective.txt) to extract and reorganize PRD tasks. This creates a surface for indirect prompt injection where malicious instructions embedded in these untrusted documents could influence the agent's task creation and PRD management logic. - Ingestion points: Data enters the context via
docs/design/gdd.md(GDD updates) and./.claude/session/retrospective.txt(retrospective findings). - Boundary markers: Absent; the skill does not specify the use of delimiters or instructions to ignore embedded commands within the ingested files.
- Capability inventory: The skill allows writing to project-critical files (
prd.jsonandprd_backlog.json), performing git commits, and sending notification messages (prd_reorganized) to other workers. - Sanitization: Absent; the workflow describes extraction and creation logic without validation or escaping of the content retrieved from the external sources.
Audit Metadata