The Agent Skills Directory

[PROMPT_INJECTION]: The skill is designed to ingest and process content from external documentation and source code URLs provided by the user. This creates a surface for indirect prompt injection where a malicious external source could contain instructions to subvert the agent's logic during manifest creation.
Ingestion points: The agent gathers and reads content from user-provided "official docs link" and "source-of-truth implementation link" (SKILL.md).
Boundary markers: The instructions do not specify the use of delimiters or directives to ignore instructions within the fetched external data.
Capability inventory: The agent has the capability to write implementation files and execute shell commands such as npx vitest and npm run (SKILL.md).
Sanitization: No explicit sanitization or validation of content from external URLs is mentioned in the workflow.
[COMMAND_EXECUTION]: The skill directs the agent to execute several shell commands for verification purposes, including npx vitest run, npm run docs, npm run check:diff, and npm test (SKILL.md). These commands are used to run local tests and documentation builds.
[EXTERNAL_DOWNLOADS]: The skill initiates network requests to retrieve and analyze content from external URLs provided during the intake process to map channel specifications.

nemoclaw-contributor-onboard-messaging-channel