herdr
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides examples of using the
herdrcommand-line tool to manage agent sessions and workspaces. These commands are part of the intended functionality of the skill within its designated environment. - [DYNAMIC_EXECUTION]: There is a bash snippet that pipes JSON output into a
python3one-liner (python3 -c 'import sys,json;...'). This is a standard and safe method for extracting specific fields from structured command output and does not execute untrusted logic. - [INDIRECT_PROMPT_INJECTION]: The skill describes methods for reading content from agents and panes (e.g.,
pane read). While this involves ingesting data from other agent processes, the skill provides descriptive guidance rather than exploitable automation routines that would act on untrusted data without oversight.
Audit Metadata