The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various shell commands using the yarn twenty CLI and utility tools like sed and test to perform operational tasks such as listing remotes, syncing changes, and inspecting application files.
[INDIRECT_PROMPT_INJECTION]: The skill reads external data from package.json and src/application-config.ts into the agent's context using sed commands.
Ingestion points: package.json and src/application-config.ts are read into the context to inspect scripts and configuration (found in SKILL.md).
Boundary markers: No explicit boundary markers or instructions to ignore embedded instructions are used when reading these files.
Capability inventory: The skill possesses extensive shell execution capabilities via yarn twenty, including the ability to deploy code, run tests, and modify remotes (found in SKILL.md).
Sanitization: There is no evidence of sanitization or filtering of the content read from the configuration files before it is processed by the agent.
[DATA_EXPOSURE]: The skill identifies ~/.twenty/config.json as a location where credentials and remote configurations are stored, highlighting it as a sensitive data path on the local system.

manage-app