defect-context-loader
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local markdown files within the .SS_WF directory to provide orientation to the agent.
- Ingestion points: The skill reads files at paths like
.SS_WF/Agent/CODE/{{$var[parent_ticket_id]s}}_CODE_GENERATION.mdand.SS_WF/Agent/Analysis/{{$var[parent_ticket_id]s}}_ANALYSIS_PLAN.md. - Boundary markers: No delimiters or warnings are used to separate the ingested content from the agent's instructions, increasing the risk of the agent following instructions embedded within the documentation.
- Capability inventory: The skill's functions are limited to file reading and context reporting; no high-risk capabilities like command execution or network exfiltration were detected.
- Sanitization: No sanitization, escaping, or validation steps are performed on the content retrieved from the files before it is processed.
Audit Metadata