ts-config
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze untrusted data from the user's codebase, such as package.json, existing tsconfig.json files, and directory structures, to provide configuration recommendations. This creates a surface where malicious instructions embedded in these project files could influence the agent's actions.
- Ingestion points: The fuse-ai-pilot:explore-codebase agent call in SKILL.md triggers the reading of project metadata and configuration files.
- Boundary markers: The instructions do not provide explicit delimiters or ignore instructions wrappers for the data ingested from the codebase.
- Capability inventory: The skill spawns sub-agents, queries external documentation via MCP (mcp__context7__query-docs), and generates/writes tsconfig.json files.
- Sanitization: There are no documented steps for validating or sanitizing the content retrieved from the local environment before it is used in prompt construction or decision-making.
- [PROMPT_INJECTION]: The skill uses strong, prescriptive language to mandate a specific internal architecture and workflow, overriding the agent's default operational logic.
- Evidence: The Agent Workflow (MANDATORY) section in SKILL.md requires the agent to spawn three specific sub-agents in parallel before proceeding with the primary task.
Audit Metadata