The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of ingesting and processing untrusted data.
Ingestion points: The researcher mode fetches data from the web (WebSearch/WebFetch), and the reviewer and analyzer modes process arbitrary code and repository data.
Boundary markers: The instructions do not define explicit boundary markers or delimiters to isolate untrusted data from the agent's internal instructions.
Capability inventory: The skill possesses significant capabilities including autonomous code generation (coder mode), file system operations, network access (researcher mode), and complex agent orchestration (orchestrator).
Sanitization: There is no mention of sanitization, filtering, or validation of the external content before it is processed by the agents.

sparc-methodology