openspec-archive-change

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands (mkdir, mv) and vendor CLI tools (openspec) that incorporate variables such as <name>. If these variables contain shell metacharacters (e.g., ;, &&, |) and are not sanitized by the execution environment, it could lead to arbitrary command execution on the host system.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted data from the file system and includes it in prompts sent to sub-agents via the Task tool.\n
  • Ingestion points: Reads task completion statuses from tasks.md and reads the content of delta specification files from the openspec/changes/<name>/specs/ directory.\n
  • Boundary markers: The prompt instructions for the sub-agent in Step 4 do not use explicit delimiters (like XML tags or triple quotes) or provide warnings to ignore instructions embedded within the file summaries.\n
  • Capability inventory: The skill has the ability to execute shell commands, read/write files in the local project directory, and invoke other agent skills/tools.\n
  • Sanitization: There is no evidence of content validation or sanitization for the data retrieved from tasks.md or delta specs before it is passed to the next step in the workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 11:15 AM