The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using lark-cli and python3. Specifically, the Phase 6 instructions for creating Feishu documents using --markdown '<content>' present a significant command injection risk if the generated report content (which includes untrusted external news) is not strictly sanitized or escaped before being passed to the shell.
[COMMAND_EXECUTION]: The skill uses inline Python script execution (`python3
<<'PY') to interact with the GitHub API. While the logic is defined within the skill, the interpolation of variables like owner_repo` into these scripts can be exploited if the input source is compromised.
[PROMPT_INJECTION]: The skill is highly susceptible to indirect prompt injection because its primary function is to ingest and summarize data from untrusted external sources, including Hacker News, GitHub Trending, and various social discussions.
Ingestion points: Data enters via news aggregators and direct API fetches from GitHub and other AI provider sites (specified in SKILL.md Phase 2 and 3).
Boundary markers: The skill lacks explicit instructions for the agent to use delimiters or ignore instructions embedded within the news content it processes.
Capability inventory: The agent has the capability to write files and execute network-connected CLI tools (lark-cli) which could be misused if a malicious news item influences the agent's behavior.
Sanitization: No explicit sanitization or filtering logic is provided for the ingested content before it is interpolated into the final report or CLI arguments.
[REMOTE_CODE_EXECUTION]: The skill performs automated network requests to the GitHub API (api.github.com) and other external sites to fetch content that is subsequently processed and used in command-line operations. This creates a dependency on the integrity of the remote data sources.

daily-ai-agent-aigc-top-news