tektus-onboarding
Pass
Audited by Gen Agent Trust Hub on May 26, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements an onboarding interview that captures arbitrary user input (role, company name, tech stack) and persists it to
.tektus/profile.yaml. This creates an indirect prompt injection surface (Category 8) because the configuration file is subsequently used by the system to automate skill activation and squad suggestions without explicit sanitization. - Ingestion points:
SKILL.md(Interview Blocks 1-3 viaAskUserQuestion). - Boundary markers: Absent; user input is directly mapped to YAML fields.
- Capability inventory: The agent is instructed to write files, install adapters, and activate specialized skills based on the profile content.
- Sanitization: No sanitization or validation of the user-provided strings is described in the instructions.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for environment setup, specifically the installation of IDE-specific adapters from local paths (
adapters/{ide}/) and the execution of the indexing toollens index ..
Audit Metadata