byted-recruitment-general
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from candidate profiles on external recruitment platforms. An attacker (candidate) could embed malicious instructions in their profile bio or experience sections to influence the agent's evaluation scoring or messaging behavior.
- Ingestion points: Candidate profiles and search result pages on LinkedIn, Liepin, and Maimai.
- Boundary markers: Absent; the agent is instructed to record information directly without explicit delimiters or instructions to ignore embedded commands.
- Capability inventory: The agent has capabilities to write markdown files to the local workspace (/root/.openclaw/workspace/jd-process/) and perform automated browser actions such as clicking 'Say Hello'.
- Sanitization: Absent; the skill does not instruct the agent to sanitize or filter profile content before processing.
Audit Metadata