The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The installation instructions in references/installation.md recommend piping a remote script directly into a shell: curl -sL https://raw.githubusercontent.com/hardhackerlabs/podwise-cli/main/install.sh | sh. This executes unverified code from an external repository belonging to an organization (hardhackerlabs) that is not identified as a trusted or well-known vendor.
[EXTERNAL_DOWNLOADS]: The skill references external binaries and source code for download from github.com/hardhackerlabs/podwise-cli, which are unverifiable and not pinned to specific safe versions.
[PRIVILEGE_ESCALATION]: The setup guide instructs the user to move a downloaded binary to /usr/local/bin/ and apply chmod +x, granting persistent execution permissions to third-party code in the system environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from podcast transcripts and AI summaries across multiple workflows (e.g., workflows/language-learning.md, workflows/episode-debate.md). Ingestion points: Data retrieved via podwise get (transcripts, summaries, etc.). Boundary markers: None identified; instructions do not provide delimiters to isolate the ingested data. Capability inventory: The agent can execute shell commands, perform network operations, and write markdown files. Sanitization: No evidence of escaping or filtering of the processed transcript data to prevent embedded instructions from influencing the agent.
[COMMAND_EXECUTION]: The skill executes local shell commands to verify installation and configuration, including podwise config show which displays the user's API key.

podwise