alterlab-phylogenetics
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes external bioinformatics CLI tools such as MAFFT, IQ-TREE 2, FastTree, and TrimAl using
subprocess.run(). This is found inscripts/phylogenetic_analysis.pyandSKILL.md. The skill constructs command arguments as lists rather than raw strings, which is a recommended practice to mitigate command injection risks while fulfilling its core purpose of running evolutionary analysis pipelines. - [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests untrusted sequence data from FASTA files and processes it using CLI tools. Ingestion points occur in
scripts/phylogenetic_analysis.py, which reads user-provided FASTA files to count sequences and passes the file paths to external binaries. The skill has the ability to execute shell commands for bioinformatics analysis and read/write files in thephylo_resultsdirectory. No explicit prompt boundary markers are used when describing sequence processing to the agent, and the skill relies on the input validation mechanisms of the underlying bioinformatics binaries to handle malformed data.
Audit Metadata