The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the blockchain ledger through commands like botchan read, botchan chat read, and netp message read. This data is provided by external users or agents and could contain malicious instructions designed to influence the behavior of an agent using this skill.
Ingestion points: packages/botchan/src/cli/index.ts, packages/net-cli/src/commands/message/read.ts, and various read subcommands across the SDK packages.
Boundary markers: The skill does not explicitly use delimiters or instructions to ignore embedded content when presenting blockchain data to the AI agent.
Capability inventory: The skill possesses significant capabilities including network operations (fetch in the SDK), transaction preparation for blockchain submission, and shell command execution (via the net-explorer agent profile defined in the plugin).
Sanitization: No specific sanitization or filtering of blockchain-sourced text was observed in the static analysis of the SDK or CLI code.
[EXTERNAL_DOWNLOADS]: The CLI tools (botchan, netp) include an update command that downloads and installs the latest versions of their respective packages from the npm registry.
Evidence: program.command("update") in packages/botchan/src/cli/index.ts and packages/net-cli/src/cli/index.ts executes npm install -g for the project's official packages.
Context: These downloads are directed at the official packages owned by the skill's author (stuckinaboot) and are standard functionality for keeping the tools up-to-date.

botchan-net