Skills
skills/hamsterider-m/personal-skills/feishu-work-archive/Gen Agent Trust Hub

feishu-work-archive

Fail

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: HIGHCREDENTIALS_UNSAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The file archive.js contains a hardcoded appId ('cli_a75960290439500d') and appSecret ('tBd5CYrc3jFTa4yQ9aD3whBMyMvDfl2w'). These are sensitive credentials for a Feishu application and should not be stored in plain text in source files.
  • [COMMAND_EXECUTION]: Multiple scripts, including agent-export.sh, export-all.sh, and test-10.sh, execute shell commands to automate browser actions and manage the local filesystem, including operations in the user's home and cloud storage directories.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface (Category 8). 1. Ingestion points: Document content is retrieved from Feishu via URLs in auto-start-export.mjs and persistent-export.mjs. 2. Boundary markers: Absent. External content is concatenated directly with frontmatter and saved. 3. Capability inventory: Filesystem write operations and browser automation via Playwright and agent-browser. 4. Sanitization: Absent. Data from external sources is processed and written to disk without validation or escaping.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 7, 2026, 02:53 AM