generate-standup
Pass
Audited by Gen Agent Trust Hub on May 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell scripts (get-standup-context.sh and post-standup-comment.sh) which invoke the 'gh' CLI and 'jq' for data retrieval and submission.
- [DATA_EXFILTRATION]: Local agent session summaries and activity logs are processed and transmitted to GitHub via discussion comments. This data transfer is the intended purpose of the skill and occurs only after user approval.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface by processing untrusted data from GitHub PRs and issues. Ingestion points: Data is ingested via the GitHub API (scripts/get-standup-context.sh) and SQL queries to the session store (SKILL.md). Boundary markers: Instructions lack explicit delimiters or guidance to ignore potential instructions within the fetched activity data. Capability inventory: The skill performs GitHub API write operations and executes shell commands (scripts/get-standup-context.sh, scripts/post-standup-comment.sh). Sanitization: JSON construction is handled via 'jq', providing protection against basic command injection in the API payload.
Audit Metadata