bmad-agent-po
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various shell commands to perform ticketing operations. These include using the GitHub CLI (
gh), Atlassian CLI (acli), anduv runto execute local Python scripts (tickets.py,read_store.py). These commands are standard for the skill's purpose of synchronizing local ticket states with remote trackers. - [EXTERNAL_DOWNLOADS]: Setup instructions in
config/jira-ticketing.tomlprovide commands to install the official Atlassian CLI (acli) using Homebrew (brew install acli). This reference targets a well-known service and is part of the tool's legitimate configuration workflow. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input, such as PRDs, briefs, and transcripts, to generate requirement tickets.
- Ingestion points:
references/spec.mdspecifies reading input from user-provided file paths or pasted content. - Boundary markers: The skill enforces structured output using markdown templates found in the
assets/directory (e.g.,ticket-template.md). - Capability inventory: The skill possesses the capability to write files and execute network-connected CLI tools for ticketing platforms.
- Sanitization: The skill implements a 'Self-validate' step in
references/spec.mdto check for coherence and ensure that no load-bearing claims are lost or misrepresented. - [OBFUSCATION]: Static analysis identified non-standard Unicode characters (such as U+2028 Line Separator) in
scripts/tests/test_tickets.py. These are utilized as test cases to verify that thetickets.pyscript correctly parses and escapes special characters during data processing, which is a standard engineering practice.
Audit Metadata