duoduo-channel-admin

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of the duoduo CLI to manage channel lifecycles and probe system state. Commands include duoduo daemon status, duoduo channel list, and duoduo channel <kind> start/stop/logs.
  • [DYNAMIC_EXECUTION]: The provided maintenance script scripts/reset-feishu-session.sh utilizes embedded Python code (via python3 -c and heredocs) to parse JSON session data and perform atomic file updates. Additionally, a run() function containing eval "$*" is defined in the script, although it is not currently invoked by the script's main logic.
  • [INDIRECT_PROMPT_INJECTION]: The skill manages descriptor.md and kind-level configuration files that contain Markdown prompts. These prompts are interpolated into the agent's identity at runtime to define its behavior for specific channels.
  • Ingestion points: Reads and modifies Markdown files (descriptors) in ~/.aladuo/var/channels/ and kernel/config/ using the patch_markdown_frontmatter.py script.
  • Boundary markers: The identity prompt is assembled by concatenating various Markdown sections without specific ignore-instructions or boundary delimiters.
  • Capability inventory: The skill possesses capabilities to install software plugins, execute shell scripts, and restart background daemon processes.
  • Sanitization: The skill preserves and applies Markdown bodies from descriptor files as authoritative prompt material without sanitization.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 12:53 AM
Security Audit — agent-trust-hub — duoduo-channel-admin