jira
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install the
aclitool from Atlassian's official site and themdadftool from the author's GitHub repository. - [REMOTE_CODE_EXECUTION]: Installation instructions for the
mdadftool include downloading a binary archive viacurl, extracting it, and moving it to a system path. - [COMMAND_EXECUTION]: The skill constructs and executes shell commands for
acliandmdadfusing user-provided input. It utilizes process substitution and temporary files for handling rich text content. - [PROMPT_INJECTION]: The skill ingests untrusted user Markdown to generate Jira ticket content. Ingestion points include summaries and descriptions; boundary markers include HEREDOCS; capabilities include work item mutations; and explicit sanitization of shell-interpreted strings is absent.
- [CREDENTIALS_UNSAFE]: The skill reads project-specific settings from
.jira-skill.jsonand global settings from~/.config/jira-skill/config.json. These files contain project keys for resolution and do not store sensitive credentials or API tokens.
Audit Metadata