codex-session-recovery
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection due to its requirement to process untrusted data from external sources.
- Ingestion points: The agent is instructed to read live tmux pane output, current process commands, and session metadata stored in JSONL files within
~/.codex/sessions/(SKILL.md). - Boundary markers: There are no explicit instructions or delimiters provided to ensure the agent ignores potentially malicious instructions embedded within the recovered session logs.
- Capability inventory: The skill has the capability to execute
tmuxmanagement commands andcodex resumecommands based on findings from the logs (SKILL.md). - Sanitization: No explicit sanitization, validation, or escaping of the session metadata is defined in the workflow.
Audit Metadata