fablecodex-workflow
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to execute arbitrary shell commands for project verification, such as running tests (
pytest) or database migrations (psql), via the--verify-cmdparameter in thecodex-fable5tool. - [EXTERNAL_DOWNLOADS]: The installation instructions require downloading a plugin from an external GitHub repository (
baskduf/FableCodex). While this is the intended distribution method, it involves fetching code from a repository outside of the platform's immediate control. - [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection because it processes untrusted external data.
- Ingestion points: The skill ingests data as 'evidence' and 'findings' via command-line arguments (e.g.,
findings add --evidence "..."). - Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands within the processed evidence data.
- Capability inventory: The skill possesses the capability to execute shell commands through the verification gate logic.
- Sanitization: No sanitization or validation of the external input is performed before it is integrated into the workflow state.
- [REMOTE_CODE_EXECUTION]: The skill facilitates the execution of code provided during the workflow's verification steps, which could be exploited if the source of the verification commands or the evidence being processed is malicious.
Audit Metadata