The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses a substantial attack surface for indirect prompt injection due to its primary function of ingesting and processing data from numerous external and attacker-controllable platforms. \n
Ingestion points: Data enters the agent context from platforms such as Twitter, Reddit, GitHub, and the general web through commands like exa.web_search_exa, twitter search, and the Jina Reader service (SKILL.md). \n
Boundary markers: There are no instructions to use delimiters or specific safety prompts to distinguish between the skill's operational instructions and the potentially malicious data retrieved from the internet (SKILL.md). \n
Capability inventory: The skill utilizes shell execution for tools like curl, gh, yt-dlp, and agent-reach, which could be targeted by instructions embedded in fetched content (SKILL.md). \n
Sanitization: No sanitization, filtering, or validation of retrieved content is mentioned or implemented before the data is processed by the agent (SKILL.md). \n- [EXTERNAL_DOWNLOADS]: The skill references well-known services and the author's repository for core functionality and documentation. \n
Uses r.jina.ai, a well-known technology service, for converting web pages into markdown format. \n
Fetches update and installation guides from the official repository at github.com/Panniantong/agent-reach. \n- [COMMAND_EXECUTION]: The skill relies on local CLI tools for environment verification and performing research tasks. \n
Executes commands such as agent-reach doctor, agent-reach check-update, and various platform-specific binaries for data retrieval. \n
Appropriately manages data by using /tmp/ for temporary output and ~/.agent-reach/ for persistent configuration storage.

agent-reach