workflow-planned-work
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions in
SKILL.mdinclude a command for the agent or user to execute a local Node.js script (node .claude/scripts/lib/workflow-skill-contract.cjs --write) to regenerate documentation blocks. This represents execution of code on the host system. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, such as product ideas, user briefs, and existing specifications, which serves as a surface for indirect prompt injection.
- Ingestion points: Data enters the context via user prompts, tickets, and existing specification files as described in
SKILL.mdandreferences/mode-idea-to-tasks.md. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the data being processed.
- Capability inventory: The skill has significant capabilities including file system writes to multiple directories (
team-artifacts/,tmp/reports/), local script execution, and the ability to spawn sub-agents (spawn_agent). - Sanitization: There is no mention of sanitization or validation of the ingested data before it is used to drive the workflow or generate artifacts.
Audit Metadata