The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the X/Twitter platform via a third-party API service.\n
Ingestion points: The fetch_tweets function in scripts/twitter_monitor.py retrieves tweet content from a remote API.\n
Boundary markers: The instructions do not define delimiters or specific warnings for the agent to ignore instructions that may be embedded within the tweet data.\n
Capability inventory: The skill has capabilities to write to the local file system, perform network requests, and execute shell commands for data synchronization and scheduling.\n
Sanitization: The script processes and stores tweet text without sanitization, creating a surface for potential downstream injection attacks.\n- [PERSISTENCE]: The skill includes instructions to establish recurring execution of monitoring tasks using system tools such as cron and launchd.\n- [COMMAND_EXECUTION]: The skill leverages shell command execution to interface with external utilities like feishu-cli for data export and management of scheduled tasks.\n- [EXTERNAL_DOWNLOADS]: The skill documentation suggests the installation of third-party tools like feishu-cli and OpenClaw to support extended functionality like Feishu synchronization and agent-managed scheduling.\n- [DATA_EXFILTRATION]: The Python script performs network operations to api.twitterapi.io to fetch data. This involves transmitting a user-provided API key to a third-party domain that is not on the whitelisted service list.

twitter-monitor