The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and executes shell commands using user-provided query strings.
Evidence: SKILL.md and resources/execution-protocol.md define the use of gh search code "<query>" and glab api "/search?scope=blobs&search=<query>" to perform code searches. While these use quoting, they represent a potential surface for command injection if the underlying platform does not provide robust sanitization of the <query> variable.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection attacks because it processes untrusted data from external sources.
Ingestion points: resources/execution-protocol.md describes gathering results from web searches, documentation libraries (via Context7), and public code repositories (GitHub/GitLab).
Capability inventory: SKILL.md specifies capabilities in the PROCESS and NETWORK scopes, including subprocess execution and network operations.
Boundary markers: The protocol lacks documented use of delimiters or instructions to the agent to ignore embedded commands within search results.
Sanitization: No explicit sanitization or filtering of the retrieved content is mentioned before it is presented to the agent's context.

oma-search