The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references an official installer at https://cdn2.flowus.cn/flowus-cli/install. It mandates that the agent must download the file, verify its SHA-256 hash, and obtain explicit user approval before execution.\n- [COMMAND_EXECUTION]: The skill extensively uses the flowus CLI to interact with the FlowUs API. It instructs the agent to use machine-readable formats and verify command success before proceeding.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content from FlowUs pages and databases, creating an ingestion surface.\n
Ingestion points: External data enters the context via flowus markdown get and flowus --json page get as described in SKILL.md.\n
Boundary markers: No specific delimiters or warnings for processed content are defined in the instructions.\n
Capability inventory: The agent has broad capabilities to create, modify, and search content, as well as perform raw API calls via flowus api call.\n
Sanitization: There are no explicit instructions to sanitize or escape retrieved content before interpolation into prompts.\n- [CREDENTIALS_UNSAFE]: The skill manages a FLOWUS_TOKEN and local session credentials. It includes proactive safeguards, such as instructing the agent to redact credentials from output and avoiding passing tokens as command-line arguments to prevent leakage via process lists.

flowus-cli