The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The scripts/bootstrap.mjs script retrieves package versioning and metadata from an external domain (https://animade.codyx.lol/api/v1/bootstrap) and uses this data to programmatically install code via npm install --global. This allows a remote service to dictate which executable code is deployed and run on the host machine.
[EXTERNAL_DOWNLOADS]: The skill initiates network requests to animade.codyx.lol to fetch installation configurations and triggers NPM to download external packages.
[COMMAND_EXECUTION]: The bootstrap script utilizes execFileSync to execute npm commands and the resulting animade-cli binary. The SKILL.md instructions further guide the agent to execute multiple shell commands for project management and onboarding.
[PERSISTENCE]: The automatic bootstrap process installs the animade-cli package globally, creating persistent executables in the system's global path that remain active across sessions.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it ingests untrusted data from external sources and local files.
Ingestion points: Data retrieved via animade-cli recipes get and animade-cli knowledge get commands, and local files like brief.md in SKILL.md.
Boundary markers: The skill does not define specific delimiters or instructions for the agent to ignore potentially malicious commands embedded in these data sources.
Capability inventory: The skill has access to shell execution via execFileSync and the CLI, as well as project modification capabilities.
Sanitization: There is no evidence of sanitization or strict schema validation for the data ingested from remote recipes or local project files.

animade