The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the asm CLI to interact with the local environment and a remote catalog. Specific commands include command -v asm for prerequisite checking, asm search for discovery, and asm bundle show for manifest validation. It also executes date -u +%Y-%m-%dT%H:%M:%S.000Z to generate ISO-8601 timestamps for the bundle metadata.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data retrieved from an external, live catalog which is not under the immediate control of the skill itself.
Ingestion points: In SKILL.md (Step 3) and references/catalog-discovery.md, the agent is instructed to read JSON output from asm search, specifically evaluating the description and installCommand fields of candidate skills.
Boundary markers: No explicit boundary markers or instructions to ignore instructions embedded within catalog descriptions are provided.
Capability inventory: The skill possesses the capability to execute shell commands (asm), write JSON files to the filesystem, and provide the user with a command to install the discovered remote code.
Sanitization: There is no evidence of sanitization or filtering of the text content retrieved from the catalog before it is used to generate the sequenced plan for the user.
[EXTERNAL_DOWNLOADS]: The skill performs network operations via the asm tool to fetch catalog data and facilitates the installation of remote skills using the asm bundle install command.

find-me-skills